Development and Testing of a New Ground Measurement Tool to Assist in Forest GIS Surveys

Round  1

Reviewer 1 Report

After realized  many changes and modifications I´m now satisfied with current quality of assessed scientific work. I have only a few minor recommendations before publication.

In Abstract (row 18) explain abbreviation RMSE (is used before abbreviation rRMSE). Append rRMSE too by positioning accuracy (not only by diameter and height).

By coordinates of Research area (row 90) I recommend use only grad and minutes, without seconds, or append range 4 boundary points (North, East, South and West) with seconds.

Row 96 – I recommend use following text: 45 x 45 m (without first "m").

Row 112 –explain abbreviation "BDS/GPS".

Check text on Figure 4. I think point A4 looks like A1. Make correct or make text larger.

Row 188. Explain symbol “ʎi” used in equation. What is (ha) in row 190? Correct this.

Row 202. I think that formula (4) is missing or not correct published.

Please check text and format text in chapter 2.5.2 (rows 266-275). Latin names with scientific abbreviation, correct is (Ginkgo biloba L.), retry this text format for all used species.

Check Figure 8. Append title of x- and y- Axis. And too for both y- axis (I think right is share or proportion, what is left?). I suppose X-axis represents iterations.

Row 578 – check and correct format of references 17., append authors of this title. Too reference 21., 32.and 45.

Author Response

I am very grateful to the experts for their comments.In the resubmitted manuscript, I have made modifications according to the comments of experts and used the revision mode to highlight where the original manuscript was modified.

Point 1:In Abstract (row 18) explain abbreviation RMSE (is used before abbreviation rRMSE). Append rRMSE too by positioning accuracy (not only by diameter and height).

Response 1: A modification was made in the revised manuscript to increase the rRMSE of the tree position.

Point 2:By coordinates of Research area (row 90) I recommend use only grad and minutes, without seconds, or append range 4 boundary points (North, East, South and West) with seconds.

Response 2: Only grad and minutes, without seconds were used in the revised manuscript.

Point 3:Row 96 – I recommend use following text: 45 x 45 m (without first "m").

Response 3: In the revised manuscript, "45m x 45m" was replaced by "45 x 45m".

Point 4:Row 112 –explain abbreviation "BDS/GPS".

Response 4: "BeiDou navigation satellite system/global positioning system" was added in the resubmitted manuscript.

Point 5:Check text on Figure 4. I think point A4 looks like A1. Make correct or make text larger.

Response 5: In the revised manuscript, I replaced it with a clearer picture and made a correction.

Point 6:Row 188. Explain symbol “ʎi” used in equation. What is (ha) in row 190? Correct this.

Response 6: “ʎi” stands for proportionality coefficient. "ha" has been changed to "h" and represents the vertical difference between the optical axis of the camera and the laser beam of the pointer.

Point 7:Row 202. I think that formula (4) is missing or not correct published.

Response 7: "formula (4)" has been added to the resubmitted manuscript.

Point 8:Please check text and format text in chapter 2.5.2 (rows 266-275). Latin names with scientific abbreviation, correct is (Ginkgo biloba L.), retry this text format for all used species.

Response 8: The English name and Latin name of the tree species were re-examined in the new manuscript and the format was corrected for all species used.

Point 9:Check Figure 8. Append title of x- and y- Axis. And too for both y- axis (I think right is share or proportion, what is left?). I suppose X-axis represents iterations.

Response 9: The x-axis represents the number of iterations. The left y-axis represents the loss. The right y-axis represents the accuracy. Titles for the x and y axes have been added to the new manuscript.

Point 10:Row 578 – check and correct format of references 17., append authors of this title. Too reference 21., 32.and 45.

References 17, 21, 32 and 35 were corrected in the revised manuscript and other references were checked.

Author Response File: Author Response.docx

Reviewer 2 Report

Forests-543903

Dear Editor,

Thank you for inviting me to review the manuscript entitled “Development and testing of a new ground measurement tool to assist in forest GIS surveys” for Forests journal.

This paper addresses an interesting and up-to-date subject. The manuscript could be improved in some places, but it adheres to Forests journal policies and topics. Below I have inserted some recommendations through which the scientific value and readability of the manuscript could be improved.

            Minor/major comments for the authors

- The ‘Abstract’ should begin with something with a phrase that explains the purpose of the paper. This current abstract starts to abruptly. I recommend a short sentence to set the tone of the following paper/research (ex. In present forest surveys, some problems occur because of the cost and time required when using external tools to acquire tree measurement, so that it is of great importance to develop a new cost-saving and time-saving ground measurement method implemented in a forest GIS survey).

R17 “ground measurement tool” repetition, please rephrase

R18 “0.222 m and 0.229m, respectively” on X and Y axes? What coordinates system? Please explain a little

R36-38 The phrase has no sense, please rephrase

R49 I strongly recommend you change “discriminates” to “differentiate”. Discriminates also has a negative meaning, you should change/replace this word throughout the paper.

R134 and 136 “this paper obtained” and “In this paper” please rephrase, ex. “using this method”

R157-158 “and ℎ represents the displacement between” it is vertical difference. Displacement is an error.

R162-163 “aiming at the target tree to be measured, the laser point is projected onto the tree to be measured.” You measure the distance between your instrument and the tree using an electronic distance measurement tool, ex. distometer. Please rephrase this idea, other persons will not understand what you are trying to explain.

R202-203 Formula 4 and 5 are missing

R215 Use DBH in formula, not D. D stands for distance

The whole “Measure DBH and standing tree height” subchapter uses notions and principles of survey engineering, trigonometric levelling, photogrammetry etc. You should cite other previous articles with similar topics. There are plenty.

R229 Try to change the colour of the text inside the pictures, it has very poor visibility

R262 Replace “shooting” with “photographing”

R307-308 “The data of tree position measured by Leica Viva GNSS-RTK system GS15 was used as the reference value of tree position in this paper.” How did you manage to obtain a fixed position using GNSS RTK method inside a forest? Did you measure the position of every tree or the position of your instrument (mobile phone, EDM tool, tripod)?

            R327-328 After using "Beidou Compation M1" to obtain the position of the high precision observation site, the coordinates were compared with the tree position coordinate estimation with the data (no estimation, GNSS RTK is the precise method) obtained with the Leica Viva GNSS-RTK.

            R344 Please rephrase, it makes no sense

            R414 I recommend you replace “this article’s mode” with “the presented method” or similar.

            R425 “after taking the required photo is taken” rephrase

Finally, I would like to congratulate the authors for the effort and scope of the article. It presents an interesting method/topic and has high readability and an interdisciplinary approach. Some aspects need to improved (ex. Abstract and Materials and methods), minor spelling mistakes and some parts need further explaining.

Author Response

I am very grateful to the experts for their comments. In the resubmitted manuscript, I have made modifications according to the comments of experts and used the revision mode to highlight where the original manuscript was modified.

Point 1:- The ‘Abstract’ should begin with something with a phrase that explains the purpose of the paper. This current abstract starts to abruptly. I recommend a short sentence to set the tone of the following paper/research (ex. In present forest surveys, some problems occur because of the cost and time required when using external tools to acquire tree measurement, so that it is of great importance to develop a new cost-saving and time-saving ground measurement method implemented in a forest GIS survey).

Response 1: Thanks to the experts for their comments, the original manuscript was revised according to the opinions of the experts. Corrections were made to the beginning of the abstract in the new manuscript and the comments from the experts were taken.

Point 2:R17 “ground measurement tool” repetition, please rephrase

Response 2: Re-expressed in the new manuscript as the following sentence. “This paper selected two plots to verify the accuracy of the ground measurement tool.”

Point 3:R18 “0.222 m and 0.229m, respectively” on X and Y axes? What coordinates system? Please explain a little

Response 3: 0.222m and 0.229m represent the RMSE of the position of the trees in the two plots, respectively, representing the error of the plane coordinates. After calculating the RMSE of the x-axis and the y-axis, respectively, the RMSE of the plane coordinates is calculated according to the formula . The WGS-84 coordinate system was chosen in this paper.

Point 4:R36-38 The phrase has no sense, please rephrase

Response 4: The new manuscript has been modified to remove the meaningless parts. Only the following sentences were retained. “Combine GIS software based on field maps with electronic devices (such as mobile phones or tablets) for forest GIS surveys”.

Point 5:R49 I strongly recommend you change “discriminates” to “differentiate”. Discriminates also has a negative meaning, you should change/replace this word throughout the paper.

Response 5: In the new manuscript, all "discriminates" were changed to "differentiate" and the relevant parts of the whole paper were modified.

Point 6:R134 and 136 “this paper obtained” and “In this paper” please rephrase, ex. “using this method”

Response 6: Changes have been made to R134 and 136 in the new manuscript, replace “this paper obtained” with “using this method”.

Point 7:R157-158 “and ℎ represents the displacement between” it is vertical difference. Displacement is an error.

Response 7: In the new script, "difference" is used instead of "Displacement".

Point 8:R162-163 “aiming at the target tree to be measured, the laser point is projected onto the tree to be measured.” You measure the distance between your instrument and the tree using an electronic distance measurement tool, ex. distometer. Please rephrase this idea, other persons will not understand what you are trying to explain.

Response 8: Modifications have been made in the new manuscript.

Point 9:R202-203 Formula 4 and 5 are missing

Response 9: Formulas 4 and 5 have been added to the new manuscript.

Point 10:R215 Use DBH in formula, not D. D stands for distance

Response 10: In the new manuscript, "DBH" was used instead of "D".

Point 11:The whole “Measure DBH and standing tree height” subchapter uses notions and principles of survey engineering, trigonometric levelling, photogrammetry etc. You should cite other previous articles with similar topics. There are plenty.

Response 11: I added references to articles on similar topics in new manuscripts, such as references 31, 32, 33, 34, 35.

Point 12:R229 Try to change the colour of the text inside the pictures, it has very poor visibility

Response 12: In the new manuscript, I changed Figure 5 to a picture with a clearer resolution.

Point 13:R262 Replace “shooting” with “photographing”

Response 13: Replace “shooting” with “photographing” in the new manuscript.

Point 14:R307-308 “The data of tree position measured by Leica Viva GNSS-RTK system GS15 was used as the reference value of tree position in this paper.” How did you manage to obtain a fixed position using GNSS RTK method inside a forest? Did you measure the position of every tree or the position of your instrument (mobile phone, EDM tool, tripod)?

Response 14: This study was carried out in the artificial forest of the Haidian District(40°1′N, 116°20′E) of Beijing, China (Figure 1). First, we chose to calibrate two temporary plots in the artificial forest. Ten sites with good signal were selected near each plot, and GNSS-RTK was used to obtain the point coordinates of the sites in the WGS-84 coordinate system. By setting up the total station at the sites, changing the site observation (moving total station observation), finally the point coordinates of each sample tree in the plot were obtained as reference values.

Point 15: R327-328 After using "Beidou Compation M1" to obtain the position of the high precision observation site, the coordinates were compared with the tree position coordinate estimation with the data (no estimation, GNSS RTK is the precise method) obtained with the Leica Viva GNSS-RTK.

Response 15: the measurement reference value of the tree height and the DBH was expressed by "estimation", "estimation" was used in this sentence. Corrections have been made in the new manuscript and the sentence has been changed to “After using "Beidou Compation M1" to obtain the coordinates of the high precision observation site, the coordinates were compared with the tree position obtained with the Leica Viva GNSS-RTK.”.

Point 16:  R344 Please rephrase, it makes no sense

Response 16: In the new manuscript, R344-366 was deleted, and a new sentence was added - "The estimated values of tree height and DBH obtained in the two plots were compared with the reference values respectively."

Point 17: R414 I recommend you replace “this article’s mode” with “the presented method” or similar.

Response 17: In the new manuscript, "this article's mode" was changed to "the presented method".

Point 18:R425 “after taking the required photo is taken” rephrase

Response 18: The new manuscript has changed the R425 to "the DBH and tree height are estimated quickly just after the tree photo is taken and in the paper estimates are compared with measurements"

Author Response File: Author Response.docx

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round  1

Reviewer 1 Report

Dear authors,

I am aware of a previous submission connected with the material of this paper but, due to time limitations, I could not make a direct comparison.

The general comment to this paper is, in the end, still negative.

The search for integrated approaches and instruments in forest mensuration and inventory practices is of great interest, hence the work has, in theory, a good potential.

The paper is still un-readable in many points.

The presentation of the new instrument requires, first of all, explicitly considering the different proposals that have already been presented and used. From my point of view there is no “traditional work mode”: Bluetooth has already been used to automatically register dbh, height, position and other measurements and dedicated stations like FieldMap are on the market since more than a decade.

Having taken into consideration existing proposals, it is necessary to let the reader understand what is in the new proposal, just a photo of it is not enough. What is the “intelligent terminal  module  based  on  the  Android  7.0  operating  system”? How could a reader reproduce your results? Are you using a hardware that is or will be on the market? If not what are the specifications? Is the software you used proprietary code that can be be acquired or are you making it available as open-source material?

The explanation of the computations applied to derive tree dbh and height measures is not at all convincing: you can not reverse the collinearity equations to compute 3D coordinates given the coordinates of a point on the photo, some auxiliary information is needed (and ‘f’, the focal length, is not enough).

To compute the topographic coordinates of the trees, distances, azimuth and bearing from observation point are required, how are these data acquired? through the processing of the single photo?

The ‘computer assisted’ species  identification procedure is potentially interesting while, as it is hidden behind specialised technical details, the presentation becomes almost pointless. How does it work in practice? Does the software expect to have a photo of a fresh leaf with a white background? Also in this field existing botanical species identification apps should be explicitly discussed.

In the attached files comments are distributed across the text.

A very deep revision is highly encouraged, I would personally be interested to understand what the paper does not explain.

Regards,

Comments for author File: Comments.zip

Author Response

The presentation of the new instrument requires, first of all, explicitly considering the different proposals that have already been presented and used. From my point of view there is no “traditional work mode”: Bluetooth has already been used to automatically register dbh, height, position and other measurements and dedicated stations like FieldMap are on the market since more than a decade.

Response : "FiledMap" is generally GIS software running on mobile phones, which is more professionally called "mobile GIS". It has been widely used in forest resource surveys. Field workers in the forest use mobile phones to collect forest resource data. The data collected includes DBH, height, coordinates, tree species, class number, tree number, canopy begins, canopy height, canopy diameter, perimeter breast height, tree health, etc. Among the collected data, the altimeter or total station is used to measure the height of the tree, the breast diameter caliper or the circumference to measure the breast diameter, and the mobile phone or the handheld GPS obtains the geographical position. In addition to the operation of mobile phones, the workers also need to operate the altimeter and the breast diameter caliper, which is inconsistent and time consuming. Therefore, this paper designs a ground measurement tool, which is mainly used to collect data by mobile phones. Compared with the current workflow, it saves time and labor.

Bluetooth technology is not the focus of this paper. Generally, field workers usually use mobile phones for positioning, but the positioning bias of mobile phones is 4m-10m. In this paper, in order to improve the accuracy of mobile phone positioning used by field workers, the RTK positioning module "Beidou Companion M1" is inherited from the ground measurement tool. "Beidou companion M1" transmits high-precision geographical location data to the mobile phone through bluetooth. Bluetooth technology is only the data transmission method used in this paper.

Having taken into consideration existing proposals, it is necessary to let the reader understand what is in the new proposal, just a photo of it is not enough. What is the “intelligent terminal  module  based  on  the  Android  7.0  operating  system”? How could a reader reproduce your results? Are you using a hardware that is or will be on the market? If not what are the specifications? Is the software you used proprietary code that can be be acquired or are you making it available as open-source material?

Response :  "Intelligent terminal module based on the Android 7.0 operating system" is a mobile phone whose operating system is Android 7.0. The technical parameters of the hardware modules of theground measurement tool designed in this paper were added to “Appendix A” in the new manuscript. These hardware modules are common on the market. In order to solve the problem studied in this paper, the ground measurement tool in this paper integrated these hardware organically and designed an algorithm for it. Readers can completely repeat our experiments based on the materials and algorithms designed by us, and help field workers to collect GIS data with mobile phones.

The explanation of the computations applied to derive tree dbh and height measures is not at all convincing: you can not reverse the collinearity equations to compute 3D coordinates given the coordinates of a point on the photo, some auxiliary information is needed (and ‘f’, the focal length, is not enough).

Response : In this paper, not only f focal length but also h=875px were known, which has been described in the last sentence of "2.2" of the old manuscript.

In order to make readers read more clearly and repeat the experiment of this paper, the algorithm is re-expressed in the new manuscript

In this paper, the lens center S of the camera is set as the origin. The origin coordinates of the object space coordinate system and the image space coordinate system are coincident, and the coordinate axis directions are completely coincident. The X-axis and the Y-axis of the photo coordinate system (two-dimensional coordinate system) coincide with the X-axis and Y-axis directions of the object-side space coordinate system and the image-side space coordinate system respectively. The rotation Angle of the coordinate system is 0, and the translation amount in all three directions is 0, so simplify the rotation matrix and the special form of collinear equation can be obtained according to the principle of collinear equation.

The collinear equation becomes the following form, which is a special form of the collinear equation. This formula can be used to calculate the coordinates of a single photo in the object space coordinate system. The specific algorithm is shown in the figure below.

                                                          （1） 

S–XYZ is object space coordinate system .xSy is photo coordinate system and S-xyf  is image space coordinates system. S is the photographic center(The common origin of three coordinate systems).  and  are the object coordinate points on the top and bottom of the tree, respectively.  and  are the object coordinate points on the left and right sides of the DBH, respectively, and  represents the vertical height of between the lens center and the laser pointer. The height of  was known. The reference value of  in this experiment is 875px. 

The coordinates of the laser point in photo coordinate system are（）, and the S coordinates of the camera's lens center in photo coordinate system are（）.

Because

Since  and  were known, the following formula was obtained, and then calculate the DBH and the height of the tree respectively.

The ‘computer assisted’ species  identification procedure is potentially interesting while, as it is hidden behind specialised technical details, the presentation becomes almost pointless. How does it work in practice? Does the software expect to have a photo of a fresh leaf with a white background? Also in this field existing botanical species identification apps should be explicitly discussed.

Response : This article does not capture the tree topographical coordinates from a single photo, but only obtains the DBH and height of the tree from a single photo. In this paper, mobile phones are used to directly obtain the geographical location of trees. At present, most of the methods are only through the built-in GPS positioning of mobile phones, but this mobile phone positioning has a big deviation, especially in the case of weak signal. In order to improve the positioning accuracy of the mobile phone in the forest, this paper integrates the “Beidou Companion M1” positioning module when designing the ground measurement tool.

The ‘computer assisted’ species  identification procedure is potentially interesting while, as it is hidden behind specialised technical details, the presentation becomes almost pointless. How does it work in practice? Does the software expect to have a photo of a fresh leaf with a white background? Also in this field existing botanical species identification apps should be explicitly discussed.

Response : field workers usually use their eyes to observe trees, and then combine their personal knowledge or experience to identify tree species , which may lead to the situation that they cannot distinguish tree species due to insufficient knowledge or experience. Based on MobileNets, a lightweight neural network friendly to mobile phone,a model was trained to assist field workers in identifying tree species under the above circumstance. This provides an effective reference for field workers to identify tree species and can assist in artificial identification of tree species.

In the current work, the existing method of tree identification is that field workers observe trees with their eyes, and then judge trees based on their personal knowledge or experience. Therefore, it may lead to the situation that they cannot distinguish tree species due to insufficient knowledge or experience.

In order to highlight the leaves, the background of these leaves is shown in white. In actual work, the white background is not required. When the field workers cannot identify the tree due to insufficient knowledge or experience, field workers can remove the mobile phone from the ground measurement tool, close to the tree, open the tree identification software, point the camera of the mobile phone to the blade of the tree and click to shoot Button. After that, the image can be automatically loaded into the interface, and automatically recognized by the program, then the information of the tree species can be displayed in the interface. The specific process is as follows.

The following is a comment in the pdf attachment for the reviewer, Reply point by point

Page 1:

Point 1: drawing geometric graphics data on the map，

Point 2:collecting tree measurements，

Point 3:registering data in GIS linked forms.

Response 1,2 and 3: The research background of this paper is derived from the case of “mobile GIS application in forest survey”. Mobile GIS is a GIS system supported by the mobile Internet, with mobile phones or tablet computers as terminals, combined with beidou, GPS or base station as positioning means. Field workers in the forest use mobile phones equipped with GIS system for data collection and GNSS function positioning of mobile phones to obtain geographic coordinates in real time. When using a smartphone to collect data, you need to draw the position of the tree on the map of the mobile phone. The present is a point on the map (geometry). We also need to enter the attribute data for the drawn point (tree position), such as DBH, tree height, coordinates, tree species, small class number, tree number, canopy begins, canopy height, canopy diameter, perimeter breast height, tree health and so on a series of data. These attribute data need to be collected by field workers, among which measuring DBH and tree height is the most time-consuming and laborious. Field workers usually use hypsometer or total stations to measure tree height, and calipers or diameter tapes to measure DBH. Finally, these attribute data are input into the mobile phone's GIS software.

  With regard to the research on the collection of forest resource data by mobile GIS, in order to let the distinguished experts better understand the operation process, I have provided the reference content of the following link.

https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-2-W2/165/2015/isprsannals-II-2-W2-165-2015.pdf

https://www.sciencedirect.com/science/article/pii/S221201731300073X

http://www.supergeotek.com/index.php/201602_forestapp/

https://link.springer.com/article/10.1007/s11676-011-0160-1

https://www.fs.fed.us/news/releases/us-forest-service-offers-new-digital-maps-mobile-devices

https://nfg.no/wp-content/uploads/2019/01/ForestInventory.pdf

https://www.trimble.com/Forestry/Mobile-GIS.aspx

Point 4: Highlighted: otal stations, | Comment Text: are not that common in forest
mensuration practices

Response 4：When field workers use mobile phones to collect GIS data, they often measure DBH and tree height through contact measurement methods, such as measuring DBH with calipers or diameter tapes, measuring tree height with hypsometer. Some field workers also use the total station to measure the height of the tree.

Point 5: Highlighted: scattered 42 survey tools | Comment Text: scattered is supposed to stand for 'not part of an organic instrumentation'?

Response 5：In the forest, field workers need to use mobile phones to collect GIS data. They need to operate external independent tools to measure standing trees, which belongs to the contact measurement method, such as measuring DBH with calipers or diameter tapes, measuring tree height with hypsometer or total stations, and finally recording the measured DBH and tree height into the mobile phone's GIS software. These tools, such as calipers or diameter tapes, hypsometer or total stations, are scattered relative to field workers, who need to operate these tools separately to measure standing trees. In this paper, the ground measurement tool, making full use of the mobile phone in the hands of field workers, can mainly operate the mobile phone to complete the standing tree measurement, making the work more coherent.

Point 6:Highlighted: tool to as | Comment Text: forest survey with GIS support? OR GIS surveys within forest areas?

Response 6：The forest GIS survey in this paper refers to the use of mobile GIS technology to collect forest data in the field. Field workers in the forest use a mobile phone equipped with GIS software to collect data in the forest. For example, the current relatively popular Mobile GIS technology (ESRI ArcGIS for Mobile, China's SuperMap for Mobile) and some open source Mobile GIS technology, forestry researchers have developed many applications for forestry investigation with these Mobile GIS technology, and Mobile GIS technology has been widely used in forestry investigation.

For more information on forest GIS surveys, you can refer to the following links.

https://www.fs.fed.us/news/releases/us-forest-service-offers-new-digital-maps-mobile-devices

https://nfg.no/wp-content/uploads/2019/01/ForestInventory.pdf

https://www.trimble.com/Forestry/Mobile-GIS.aspx

Page2

Point 7:Highlighted: Due to overreliance on traditional external forest measurement tools, |Comment Text: overreliance? a strange expression

Response 7：In the forest, field workers holding the mobile phone and using the mobile GIS technology to survey, it is necessary to record the tree position, tree height, DBH and other attribute information into the mobile phone GIS software. How to get the geographical location, breast diameter, tree height, tree species and other data? At present, the conventional working method is to use the GPS of the mobile phone, diameter tapes, hypsometer or total stations to obtain the position, DBH and tree height respectively. Field workers use their eyes to observe trees, and then combine their personal knowledge or experience to identify tree species. Finally, the data such as position, DBH, tree height, and tree species are input into the GIS software of the mobile phone. In this process, in addition to the need to operate the mobile phone in hand, field workers also needs to separately operate measuring tools such as a diameter tape, hypsometer or a total station. The new manuscript has been revised and re-expressed.

Point 8:Highlighted: o a certain extent, this model provides an effective 74 discriminant basis for field workers  on the ground and provides a reference for research on tree or plant  recognition in other related fields. | Comment Text: I do not understand

Response 8：In the field investigation, field workers usually use their eyes to observe trees, and then combine their personal knowledge or experience to identify tree species, which may lead to the situation that they cannot distinguish tree species due to insufficient knowledge or experience. This paper uses deep learning technology to train the tree species recognition model. When field workers can't accurately judge the tree species, they can refer to the recognition result of the model and use the tree identification method of “subjective discrimination + model recognition” to ensure the accuracy of tree species identification to a certain extent.

This article uses MobileNets to train tree species recognition models and deploys them to mobile phones. When field workers cannot identify the tree species due to insufficient personal knowledge or experience, the tree can be automatically identified by taking a picture of the leaf on the mobile phone. The tree species identification method provides an effective reference for field workers.

Point 9:Highlighted: (false acacia) | Comment Text: ???

Response 9：The English name is False Acacia or Black Locust. Latin (Robinia pseudoacacia Linn.).

Point 10:Highlighted: Materials and Methods | Comment Text: A para dedicated to reference preceding proposals for similar tool is absolutely necessary: e.g. FieldMap

Response 10：In order to more clearly describe the purpose of this paper, this article adds "FieldMap" related research in the "1.Introduction" section. You can also refer to the following links.

https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-2-W2/165/2015/isprsannals-II-2-W2-165-2015.pdf

https://www.sciencedirect.com/science/article/pii/S221201731300073X

http://www.supergeotek.com/index.php/201602_forestapp/

https://link.springer.com/article/10.1007/s11676-011-0160-1

https://www.fs.fed.us/news/releases/us-forest-service-offers-new-digital-maps-mobile-devices

https://nfg.no/wp-content/uploads/2019/01/ForestInventory.pdf

https://www.trimble.com/Forestry/Mobile-GIS.aspx

Page 3:
Point 11:Highlighted: ntelligent 93 terminal module based on the Android 7.0 operating system |Comment Text: what is this? is it the cellphone-like tool one can guessfrom the photo or is it the round box that seems to be slightly enlarged in the detailfoto? In both cases description is quite vague.

Response 11： “intelligent terminal module based on the Android 7.0 operating system”. This is a mobile phone based on the Android operating system. It is the core device of the instrument in this paper and provides the operating environment of the software in this paper. It has various sensors such as cameras, bluetooth, angle sensors and direction sensors. When field workers use smart phones running GIS software to collect forest resources data, they need to operate the smartphone to open the map and then collect data. In “Appendix A”of the resubmitted manuscript, the configuration and parameters of the mobile phone for the experiment in this paper are provided. The specific technical parameters of each part of the whole instrument (including mobile phone) are listed in detail in “Appendix A”.

Point 12:Highlighted: laser indicator | Comment Text: just a pointer or does it measure distances?

Response 12：The resubmitted manuscript is denoted as "laser pointer". The laser pointer can emit a laser beam parallel to the photographic direction (depth of field). The vertical height of the laser pointer from the center of the lens is used to obtain the proportionality coefficient of the standing tree photo. The vertical height of this experiment is 875px, which can be adjusted as needed in actual work. By means of normal case photography, the photographing direction (main optical axis) of the mobile phone lens is parallel to the horizontal plane during the photography process, and the photographed picture includes the tree top and the bottom of the tree. After taking a picture of a tree, the coordinates of the laser point can be obtained on the screen of the mobile phone (photo coordinate system). At this time, the laser indicator has the function of "pointer".

In this paper, not only f focal length but also h=875px were known

As shown in the picture above, there is no need to measure the distance from the instrument to the tree, because the vertical height h between the camera center and the laser pointer center was known. In addition, the coordinates of the camera center (the coordinates of the center point of the photo in photo coordinate system) can be obtained on the screen of the mobile phone. We can calculate the distance between the center point of the photo and the point pixel of the laser projection on the photo, and the vertical height h between the center of the laser pointer and the center of the camera is 875px(can be adjusted according to actual needs during work), and the ratio of pixel distance between h and two points can be used as a proportional coefficientλ ( is the pixel distance between the laser point and the camera center point in the vertical direction). In the work, h is input into the software, and the algorithm in this paper can be further used to calculate the DBH and tree height.

Point 13:Highlighted: photographic platform | Comment Text: more probably this il the celllinke box, isn't it?

Response 13：A photogrammetric platform is used to place and fix mobile phones. The platform can adjust the vertical height between the mobile phone lens center and the laser pointer. Keep the vertical height (h) unchanged in the process of photography, and the position of the mobile phone can be rotated 360 degrees horizontally, which is convenient for aiming at trees.

Point 14:Highlighted: mobile terminal module | Comment Text: aka 'terminal module'?

Response 14： "mobile terminal module" and "terminal module" are both "mobile phones", which are the core part of the ground measurement tool in this article, so they are called "mobile terminal module" in the old manuscript. In order to facilitate the reader to understand and repeat the experiment of this article, I refer to them as "mobile phones" in the resubmitted manuscript.

Point 15:Highlighted: raditional RTK or GPS equipment cannot directly transmit data to the terminal's GIS system and 105 | Comment Text: which is the reference: in
traditional survey tree position is not registered; other tools have already solved the
'problem

Response 15：In the mobile GIS survey, field workers usually use mobile phone positioning, but the positioning accuracy of the mobile phone itself is limited, and there is a large deviation. In order to obtain high-precision geographic location, some field workers use RTK to locate, and then input the location data into the mobile GIS software. Traditional RTK devices are also inconvenient to carry, and data cannot be directly transmitted to the mobile phone. There are two reasons for using the "Beidou Companion M1" positioning module. First, it is an RTK device that can provide high-precision position data. Compared with the traditional RTK, it has the advantages of small size and light weight(less than 70g).It can improve the problem of poor measurement accuracy caused by only relying on mobile phone positioning. Second, the positioning data of "Beidou Companion M1" directly transmits the location data to the mobile phone through bluetooth. Using bluetooth technology to transmit data is not the research focus of this paper. The focus of this paper is that the ground measurement tools organically integrate the existing RTK positioning devices on the market to improve the positioning accuracy of mobile phones in the forest. Bluetooth technology is only used as a means of communication between RTK positioning module and mobile phone. In the forest, the purpose of this paper is to use mobile phones to obtain high-precision geographical location.

(Conventional RTK equipment)

   (“Beidou companion M1”can transmit high-precision location data to mobile phones using bluetooth)

Point 16:Highlighted: When a laser designator is used in terrestrial photogrammetry, the coordinate reference frame of a 107 tree’s image is determined to ensure the
accuracy of the DBH and tree height measurements. | Comment Text: Does the
light of the laser pointer trace one of the axes of the 3D reference system built
around the photo?

Response 16：At work, the laser spot emitted by the laser pointer falls on the trunk. After taking a photo of the standing tree, the coordinates of the laser point (photo coordinate) can be obtained through finger touching and clicking on the picture (the picture is loaded on the screen of the mobile phone). The coordinates of the center point of the photograph are the coordinates of the lens center of the camera (the camera has been calibrated) and calculate the pixel distance between the two points in the vertical direction of the photo (This pixel distance corresponds to the vertical height  between the laser designator center and the lens center in the real world. In this experiment,  is 875px. Field workers can adjust the value of h as needed and input the value of  into the software when calculating.). 

In this paper, the phone's camera lens center S is set as the origin. The origin coordinates of the object space coordinate system and the image space coordinate system are coincident, and the coordinate axis directions are completely coincident. The X-axis and the Y-axis of the photo coordinate system (the coordinate system of the picture, two-dimensional(2D) coordinate system) coincide with the X-axis and Y-axis directions of the object space coordinate system and the image space coordinate system respectively. The Z-axis direction is the direction of (depth of field) photography, and the X-axis and Y-axis constitute the object space coordinate system, which is a three-dimensional(3D) coordinate system. The vertical height of the laser point and camera’s center in the Y-axis remains unchanged during photography process (the vertical height  between the laser designator and the center of the lens). The coordinates of the laser point in image space coordinates system(The image space coordinate system is a common transitional coordinate system, which is used to represent the position of the image points in the image space. In order to transform the spatial coordinates of the image points, the coordinate system describing the position of the image points in the image space should be established.) are （，, ）and in photo coordinate system are（，）. The S coordinates of the camera's lens center in image space coordinates system are （，，）and in photo coordinate system are（，）.

The coordinates of the left and right side of the DBH on the image can be obtained (on the mobile phone screen, select the coordinates by touching the point). Finally, the value of the DBH can be calculated through the formula:

D=

The  in the formula correspond to the coordinates of DBH or top and bottom of the tree in the photo coordinate system, respectively. Point  on the left side of the DBH and point  on the right side of the DBH are selected on the photo of the tree to be tested, and the corresponding coordinates in the photo coordinate system are （，）and（，）, respectively. The coordinates of the laser point and lens center in the photo coordinate system are （，）and（，）, respectively.  is the focal length of the mobile phone camera, and 、、 are the coordinates of the object space point of the tree top and the bottom of the tree to be tested.  is the vertical height between the laser pointer and the center of the lens, and its value is 35 cm. is a scale parameter（）.  

Point 17:Highlighted: photogrammetric platform | Comment Text: is this the tripod?

Response 17：The photographic surveying platform consists of a tripod and a rotating platform (a device that can be rotated). It is used to place and fix the mobile phone while working, and adjust the vertical height between the laser pointer and the mobile phone camera.

Point 18:Highlighted: designato | Comment Text: pointer (or distantiometer)?

Response 18：The laser pointer can emit a laser beam parallel to the photographic direction (depth of field). The vertical height between the laser pointer and the center of the lens is used to obtain the proportion coefficient of the standing tree photo. The vertical height of this experiment is 875px, which can be adjusted as needed.

During the photography process, The vertical height of the laser point and camera’s center in the Y-axis remains unchanged during photography process (the vertical height  between the laser designator and the center of the lens). The coordinates of the laser point in image space coordinates system(The image space coordinate system is a common transitional coordinate system, which is used to represent the position of the image points in the image space. In order to transform the spatial coordinates of the image points, the coordinate system describing the position of the image points in the image space should be established.) are （，, ）and in photo coordinate system are（，）. The S coordinates of the camera's lens center in image space coordinates system are （，，）and in photo coordinate system are（，）.

Point 19:Highlighted: This paper has developed a GIS system, named AoGIS, t | CommentText: The AoGIS system developed for this work???

Response 19：Because the research background of this paper comes from the scene of field workers using the mobile phone equipped with GIS software to collect forest resource data, this paper developed "AOGIS" (a GIS software running on mobile phone), and field workers can use the mobile phone equipped with "AOGIS" to collect forest resource data. In this experiment, the position of the obtained trees is displayed on the map, and then the measured DBH and tree height are input into the attribute table of the GIS software.

Page 5
Point 20:Highlighted: omplete | Comment Text: an image including tree top and base?

Response 20：The trees are fully imaged in the photo. (The photographs taken include the treetops and the bottom of the trunk, the left and right sides of the DBH)

Point 21:Highlighted: normal case photography | Comment Text: vague

Response 21：“Normal case photography” is a close range photogrammetry method in which the photographing direction is perpendicular to the photographing baseline direction and is in a horizontal position, that is, in the main horizontal plane. (In the measurement of this article, the shooting direction of the phone lens (main optical axis) is parallel to the horizontal plane when shooting.)

Point 22: Highlighted: During the photography process, 148 the height of the laser point andthe Y axis of the photographic center remain unchanged. The L 149 coordinates ofthe laser point on the image are (x3 , y3 ,), and the S coordinates of the camera'slens 150 center are (x0 , y0 ,). | Comment Text: I tried to interpret this geometry butI found no solution: you are mixing 2D and 3D elements

Response 22：I didn't mix 2D and 3D elements. Maybe it is because English is not my native language that I have some problems in expression. I rewrote the statement and redrew the geometry.

In this paper, the phone's camera lens center S is set as the origin. The origin coordinates of the object space coordinate system and the image space coordinate system are coincident, and the coordinate axis directions are completely coincident. The X-axis and the Y-axis of the photo coordinate system (the coordinate system of the picture, two-dimensional coordinate system) coincide with the X-axis and Y-axis directions of the object space coordinate system and the image space coordinate system respectively. The z-axis direction is the direction of (depth of field) photography, and the X-axis and Y-axis constitute the object space coordinate system, which is a three-dimensional coordinate system. During the photography process, the height of the laser point and the Y axis of the photographic center remain unchanged(the vertical height h between the laser pointer and the center of the lens). The coordinates of the laser point in image space coordinates system(The image space coordinate system is a common transitional coordinate system, which is used to represent the position of the image points in the image space. In order to transform the spatial coordinates of the image points, the coordinate system describing the position of the image points in the image space should be established.) are （，, ）and in photo coordinate system are（，）. The S coordinates of the camera's lens center in image space coordinates system are （，，）and in photo coordinate system are（，）.

Point 23:Highlighted: he image space coordinate system are (x1 , y1 , | Comment Text:what is this again?

Response 23：In order to facilitate the transformation of spatial coordinates, it is necessary to establish a coordinate system describing the position of the image points in the image space, that is, the image space coordinate system, which is a transitional coordinate system to represent the position of the image points in the image space.

The image space coordinate system takes the Photography site (or projection center) S as the coordinate origin and the principal optical axis of the camera  as the Z axis of the coordinate system. The X and Y axes of the image space coordinate system are parallel to the X and Y axes of the photo coordinate system, respectively. This coordinate system can be easily linked to the photo coordinate system. In this coordinate system, the z coordinate of each image point is equal to the length of the principal optical axis of the camera , but the sign is negative.

The DBH and the tree height to be measured can be measured consecutively, and the bottom of the trunk (the root surface position)  and the tree top  are selected on the image of the target tree.( The coordinates of  in photo coordinate system corresponding to  are （,）and the coordinates in image space coordinates system are （,,）. The coordinates of  in photo coordinate system corresponding to  are （, ）and the coordinates in image space coordinates system are （, ,）.) Assume that the coordinates of  in object space coordinate system are (, , ), and the coordinates of  and  in object space coordinate system are solved by the form of the collinear equation, and the tree height is further calculated.

For the contents of the image space coordinate system, you can see the link below.

Point 24:Highlighted: Collinearity Equation | Comment Text: From:
https://en.wikipedia.org/wiki/Collinearity_equation
"The collinearity equations are a set of two" [NOT 3] "equations, used in
photogrammetry and remote sensing to relate coordinates in a sensor plane (in two
dimensions) to object coordinates (in three dimensions)."  You seem to
compute 3D coordinates from photo coordinates!

Response 24：This paper does calculate the three-dimensional coordinate points of the vertical plane of the tree by h (875px) and focal length  in the photo coordinate system (2D coordinates) (but it does not apply to the point coordinates of other objects outside the plane of the standing tree). In more understandable words, given the actual length of a line segment on the vertical plane (the vertical height between the main optical axis and the laser point, h in this paper, is 35 cm) and focal length f, the proportionality parameters of the vertical plane projected on the photograph and the actual distance between the photography center and the tree can be calculated, and then the three-dimensional coordinates of the image point on the plane where the tree is located can be solved.( In fact, the z values of these three-dimensional coordinates are the same, because the distance (depth of field) is the same at the time of shooting.)

The vertical height between the laser point and the center of the lens was fixed and known. That is , 35 cm, in this paper. By means of normal case photography, the photographing direction (main optical axis) of the mobile phone lens is parallel to the horizontal plane during the photography process in this paper. At this time, a photo is taken at the standing tree to be measured, and the coordinates of the laser point and the center of the lens on the picture are respectively（,）and（,）, at this time , =△. The proportionality coefficient of the plane where the standing trees are located is . The left point  and the right point  of the DBH are selected, and the corresponding point coordinates on the photo are（,）and（,）(the same horizontal direction), so .

= (proportional coefficient of the plane where the standing trees), which is substituted into the following formula.

By taking （，）and（，）into the above formula, the object space coordinates of the left side and the right side of the DBH can be obtained, that is, ，，）and（，，）.( Because the assumed coordinate axes of the three coordinate systems coincide with the axis direction, and the coordinate system has no rotation and translation, there is no need to acquire 3 outer azimuth elements.)

Page 6
Point 25:Highlighted: The image space coordinate system of the terrestrial surveying tool coincides with the origin of 165 the object space coordinate system, and the
photographic center S, the image space coordinate point 166 and the object space
coordinate point are in a straight line | Comment Text: I can not follow you

Response 25：In this paper, the lens center S of the camera is set as the origin. The origin of the object space coordinate system and the image space coordinate system coincide, and the direction of the coordinate axis is completely coincident.     The X-axis and the Y-axis of the photo coordinate system (two-dimensional(2D) coordinate system) coincide with the X-axis and Y-axis directions of the object space coordinate system and the image space coordinate system, respectively. The rotation angle of the coordinate system is 0, and the translation amount in all three directions is 0, so the rotation matrix is simplified to obtain a special form of the collinear equation.

You can refer to the links below.

（https://baike.baidu.com/item/%E5%85%B1%E7%BA%BF%E6%96%B9%E7%A8%8B/2055962）

            （1） 

Formula 1 can be used to calculate a single photo. (The rotation angle of the coordinate system is 0, and the translation amount in all three directions is 0. The value of the simplified rotation matrix is 1, and a special form of the collinear equation can be derived. The collinear equation can be used not only for aerial photogrammetry, but also for terrestrial photogrammetry.)

Point 26: Highlighted: otation matrix i | Comment Text: this is not a rotation

Response 26：The rotation matrix in the  is expressed as . Because the measurement method of this paper, the rotation angle of the coordinate system is 0, and the translation amount in all three directions is 0, and the value of the matrix is 1. It says that the rotation matrix is simplified.

Point 27:Highlighted: Formula (3) | Comment Text: is not comprehensible, where do theright hand side elements come from?

Response 27：The elements on the left represent the coordinates of the points in the object space coordinate system, and  is the proportionality coefficient of the photo of the standing tree. In order to be more standardized, it has been changed to the following form.

The object space coordinates (3D) of (, , ) are obtained as the standing tree to be measured, and  represents the vertical height between laser pointer and lens center (35 cm in this experiment). and  are image plane coordinates (2D).  is the ordinate of the laser point in the photo coordinate system, and f is the focal length of the camera, the value of  was known.

Point 28:Highlighted: in the case of steep terrain, it is necessary to use a range finder 191 to obtain the distance from the instrument to the trunk, | Comment Text: From this sentence I understand that the laser you are mentioning is just a pointer, it does not measure the distance.  If so I do not understand how you compute dbh
and height:  how do you scale the distance between to points on the photo
(even using f) to the corresponding 3D real world distance? You are not using the
measure of any reference object either.

Response 28："reference objects" is used in this article. The vertical height h between the laser pointer and the center of the lens is the reference object, and the value of h was known (35 cm in this experiment). In practical work, h can be adjusted according to the specific situation, and the value of h can be input into the software. It has been explained in the last sentence of the old manuscript "2.2". In order to make the reader easier to understand, it is re-explained in the new manuscript.

The focal length  was known. The proportionality coefficient of the plane where the standing trees are located is .  are the ordinate of the laser point and the center of the photo (principal point of photograph) in the photo  coordinate system, respectively.=△. With the known focal length , the distance between the center of the lens and the trunk can be calculated(Although the distance from the instrument to the tree can be obtained by this formula, this value is not needed in this paper.).

In the process of carrying out the research work in this paper, considering the occurrence of canopy occlusion in the forest, the measurement tools in this paper may not capture the treetops, and the tree height cannot be measured. Therefore, the algorithm of formula 8 is used to measure the tree height. The way of calculating the tree height and the way of calculating the tree height of a single photo is two different algorithms for measuring the tree height. It is necessary to measure the distance using a laser rangefinder alone, and then calculate according to formula 8. Because this paper guarantees that the trees are observed by the instrument in the experiment, this method was deleted from the new manuscript in order to make the logic of this paper more clearly. Although this method is not used in the experiment in this paper, explain to the expert here and attach a software interface that uses this formula to calculate the tree height.

Page 7

Point 29:Highlighted: Deep separable convolutional network structure algorithm | Comment Text: Only few passages are comprehensible in this paragraph.

Response 29：Additional references have been added to the resubmitted manuscript to help readers better understand the MobileNets algorithm. In the forest GIS survey progress, field workers observe trees with their eyes, and then judge tree species based on personal knowledge or experience. There may be cases where the trees cannot be accurately identified due to insufficient of knowledge or experience. In this paper, the model was trained by the MobileNets. When the field workers cannot observe the tree species through personal knowledge or experience. Using a mobile phone to shoot the leaves of a tree and identify the tree species by model. The research in this paper provides an effective reference for the field workers in comparison with the fact that field workers cannot identify the tree species.

Page 8

Point 30:Highlighted: necessarily the best way. | Comment Text: could you please reference alternatives?

Response 30：In the field survey of forest GIS, the field workers identified tree species by observing trees with their eyes and discriminating tree species based on their subjective experience and knowledge.

field workers usually use their eyes to observe trees, and then combine their personal knowledge or experience to identify tree species , which may lead to the situation that they cannot distinguish tree species due to insufficient knowledge or experience. The research in this paper provides an effective reference for the field workers in comparison with the fact that field workers cannot identify the tree species.

Point 31:Highlighted: an assist in artificial identification, and provides a 251 reference for
related tree species identification research. | Comment Text: unlinked sentence
fragment

Response 31：It has been revised and redescribed in the resubmitted manuscript

When using smart phones to collect GIS data in the forest, the field workers usually observe the trees with their eyes, and then judge the tree species based on their personal knowledge or experience. As possible the workers cannot discriminate tree species, this paper uses the MobileNets algorithm to identify the tree species. For the above situations, this paper uses the method of shooting the tree leaves with a smart phone to assist the artificial discrimination, which can provide an effective reference for the workers to distinguish the tree species.

Point 32:Highlighted: Ginkgo, Eucommia bark, 254 Fraxinus chinensis Roxb, False Acacia,Platycladus orientalis, Shantung Maple, paper mulberry, Chinese 255 Red Pine,Koelreuteria paniculata, Quercus variabilis and elm. | Comment Text: why are some latin names missing?

Response 32：Latin names of some tree species have been added to the resubmitted manuscript

Page 9

Point 33:Highlighted: Figure 7. Tree leaf specimens. | Comment Text: images
a white background are expected by the system during operations?

Response 33：The white background was used to highlight the leaves. White background is not required in the actual operation recognition process. In the resubmitted manuscript, the operation flow of identifying tree species is added.

Point 34:Highlighted: Model training | Comment Text: I could find almost no comprehensible information in this paragraph

Response 34：If the reader needs to repeat the experiment of tree species recognition in this paper, this paragraph provides the reader with the training environment and technical methods, training parameters and training sample numbers required for the experiment, in order to allow the reader to repeat the experiments in this paper. The picture shows the process of 2000 iterations, including the precision of test data set, the precision of training data set and the dynamic change of loss

Page 10

Point 35:Highlighted: assisting forest GIS surveys | Comment Text: what do you mean by'assisting'?

Response 35：The " assisting " mentioned in this paper refers to the conventional mobile GIS survey. In addition to the mobile phone, the field worker also needs to carry calipers or diameter tapes, hypsometer, total station and other tools to measure the DBH and tree height, which is a contact measurement method. The ground measurement tools in this paper are highly integrated, it can measure DBH and tree height continuously by operating a smartphone, which is a non-contact measurement method and saves the physical strength and time of the worker. This paper believes that it can assist the forest GIS survey.

Point 36:Highlighted: Under | Comment Text: Using?

Response 36：Resubmitted manuscript has been modified

Point 37:Highlighted: standard | Comment Text: reference ??

Response 37：Resubmitted manuscript has been modified

Point 38:Highlighted: During the work, the baseline height of the lens center to the 298 laser indicator should be determined. | Comment Text: Why? what do you mean?

Response 38：The vertical height of the laser pointer to the center of the lens was known, that is, h in figure 4 and formula (3), in this experiment, h is 875px. This h is the object reference for the purpose of getting the scaling coefficient like the photo coordinate system. In practical application, the height value can be adjusted according to the distance of the standing tree, but it should remain unchanged in the photography process (that is, the direction of photography is parallel to the direction of laser pointer).

 We can calculate pixel distance between the center point of the photo and the laser point on the photo. The vertical height from the center of the laser pointer to the center of the camera is h. In this paper, 875px is used (it can be adjusted according to actual needs in the work). The ratio of the pixel distance between h and two points can be used as a proportional coefficient ( represents the pixel distance of the laser point and the camera center point in the vertical direction). In the work, h is input into the software, and the algorithm in this paper can be further used to calculate the DBH and tree height.

Point 39:Highlighted: were used as standard | Comment Text: have been compared with the reference values。

Response 39：This paper collects the leaves of 12 tree species and the result of artificial identification as the reference value. Use the confusion matrix and ACC to represent the test results of the model.

Page 11

Point 40:Highlighted: rom using only the mobile terminal. | Comment Text: ???

Response 40：In the resubmitted manuscript, "mobile terminal" is described again and changed to "mobile phone".

Point 41:Highlighted: was 5.684 m, the minimum bias was 4.892 m | Comment Text: no point in going down to millimeters What is 'minimum bias' do you experience a systematic error?

Response 41：In this paper, the collected coordinates were sorted out and the coordinates measured by the （Leica）Viva GNSS- RTK system were used as reference values. The coordinates measured by the RTK system were accurate to millimeters. Therefore, in order to verify the positioning accuracy of the tool in this paper, the coordinates were also accurate to millimeters.
Point 42:Highlighted: mobile terminal. | Comment Text: that is the Leica station?

Response 42：Leica Viva is used as a reference for measuring coordinates. Mobile terminal has been changed to "mobile phone"

Point 43:Highlighted: encounter | Comment Text: ?? the tree?

Response 43：The resubmitted manuscript has been revised to "When the observation point cannot observe the trees"

Point 44:Highlighted: e-established, | Comment Text: moved??

Response 44：The resubmitted manuscript has been revised to "move the instrument in this paper and select the observation site again".

Point 45:Highlighted: measuring the | Comment Text: the evaluation of the measures of

Response 45：The resubmitted manuscript has been revised to "the evaluation of the measures of".

Point 46:Highlighted: observational conditions | Comment Text: what do you mean

Response 46：In the case where the instrument of this paper can observe trees.

Point 47:Highlighted: Figure 10 | Comment Text: Figs.s 10 and  11 - The graphs are not representing the relations in the way that one would expect it. The estimated value is expected on the vertical axis (vertical displacements between points and the line usually represent the the residuals). More over, residuals (o relative residuals) should be plotted against the  predictor in order to appreciate variability and accuracy. Finally the limits of the area occupied by the dots is quite un-naturally straight ad well defined!

Response 47：The figure of the estimates and reference values were redrawn in the new manuscript.

In Fig. 10, the horizontal axis is the reference value of the diameter tape measurement, and the vertical axis is the value of DBH based on the photo solution.

In Fig. 11, the horizontal axis is the reference value measured by the total station, and the vertical axis is the value of the tree height calculated based on the image in this paper.

Page 13

Point 48:Highlighted: and confidence probability of the single image in more than 369 95% of the 91% of test datase | Comment Text: what?

Response 48：Photos whose model recognition accuracy is more than 95% account for 91% of the whole test data set

Page 14

Point 49:Highlighted: by at least | Comment Text: to just three??

Response 49：The instrument in this paper is developed and designed for mobile GIS collection of forest resources. The field workers operate mobile phones to collect forest resources, and need to operate other independent equipment to measure the DBH and tree height of the standing trees. For example, use a hypsometer or a total station to measure the height of the tree, and use a breast caliper or a square to measure the diameter of the DBH. The ground measurement tool in this paper has the functions of measuring tree height, DBH and high-precision positioning. It is mainly able to complete the measurement of the standing tree by operating the mobile phone, and there is no need to operate other tools.

Point 50:Highlighted: RMB | Comment Text: Is this a currency?

Response 50：This is the currency of the People's Republic of China, abbreviated as RMB or CNY. For the convenience of reading, this article is uniformly expressed in US dollars.

Point 51:Highlighted: Ground measurement tool | Comment Text: In the photo I spot threepieces, beside the tripod: the cell-phone-camera, the laser pointer and the round white box。

Response 51：The instrument in this paper is developed and designed for mobile GIS collection of forest resources. The field workers operate mobile phones to collect forest resources, and need to operate other independent equipment to measure the DBH and tree height of the standing trees. For example, use a hypsometer or a total station to measure the height of the tree, and use a breast caliper or a square to measure the diameter of the DBH. The ground measurement tool in this paper has the functions of measuring tree height, DBH and high-precision positioning. It is mainly able to complete the measurement of the standing tree by operating the mobile phone, and there is no need to operate other tools.

The instrument in this paper is composed of various parts and is an organic whole. The cost of the whole instrument is no more than 3500RMB and the weight is no more than 2kg, which is very convenient to carry. In the actual GIS survey, the working mode belongs to the non-contact measurement of DBH and tree height, which is different from the use of DBH caliper and hypsometer, saving the physical strength and time of the field workers.

Page 15

Point 52:Highlighted: o a certain extent. | Comment Text: what do you mean?

Response 52：In the field investigation, field workers usually observe the trees through the eyes, and judge the species of trees based on their personal knowledge or experience. Considering that field workers may not be able to effectively identify the tree species due to insufficient knowledge or experience, the data can not be collected effectively. This paper uses the MobileNets algorithm to identify the tree species, and then deploys the trained model on the mobile phone to identify tree species by taking pictures of leaves on mobile phones. The use of deep learning to assist artificial recognition can provide an effective reference for field workers.

Point 53:Highlighted: shielded | Comment Text:??

Response 53：Changes have been made in the resubmitted manuscript: high density forests or mutually screened forests

Point 54:Highlighted: Although this equipment is a new choice for forest GIS field surveys, it still has some limitations. | Comment Text: re-phrase

Response 54：To discuss the technical limitations of this article. Changes have been made in the resubmitted manuscript

Point 55:Highlighted: stand | Comment Text:??

Response 55：Resubmitted manuscript revised to “standing”

Point 56:Highlighted: or applying the method to a place where the 433 appropriate sparse trees are blocked | Comment Text: ????

Response 56：Apply observation points where tree density is appropriate.

Point 57: Highlighted: when the tree height is too small (less than 5 m), the instrument is 443 unsuitable for use | Comment Text: why?

Response 57：Since the tree height measured in this paper ranges from 5.54 m to 19.82 m, trees that are too high or too low will cause the ground measurement tool to fail to capture the full image of the standing timber. A change has been made in the resubmitted manuscript.

Point 58: Highlighted: needs further study in this paper | Comment Text: ?? what do you mean?

Response 58：The tree species identification model in this paper can only identify 12 tree species in the training sample. For more, different species cannot be identified because they are not trained. The addition of training samples to identify more species is a further study.

Point 59:Highlighted: he most important reason for using the MobileNets algorithm for identifying tree 452 species is when a field worker cannot accurately identify trees by employing naked eye recognition 453 or for other subjective reasons | Comment
Text: this is not a sentence

Response 59：The resubmitted manuscript has been revised and expressed

Page 16

Point 60:Highlighted: it is more convenient to obtain tree height, DBH and tree position, and
it saves 463 investigative cost and time. | Comment Text: ???

Response 60：The instrument in this paper is developed and designed for mobile GIS collection of forest resources. The field workers operate mobile phones to collect forest resources, and need to operate other independent equipment to measure the DBH and tree height of the standing trees. For example, use a hypsometer or a total station to measure the height of the tree, and use a breast caliper or a square to measure the diameter of the DBH. This kind of measurement belongs to contact measurement. The ground measurement tool in this paper has the functions of measuring tree height, DBH and high-precision positioning. It is mainly able to complete the measurement of the standing tree by operating the mobile phone, and there is no need to operate other tools. The measurement method is non-contact measurement.

Author Response File: Author Response.docx

Reviewer 2 Report

Contribution of same authors was on base of reviewers comments was rejected, significantly changed and modified. In spite of my statements in the previous review, now I have suggestion to the level of English. The abstract would need to be revised (revised from native speaker).  I think it is no good three times repeat… are used to obtain ... "in the first sentence. Introduction was modified and abbreviations were explained, it is OK. Instead term “altimeter” I would use hypsometer or height meter. It was not measure altitude, but the tree height. Purposes of your work would I would have once modified or specify, the goal is to present a new cost-saving technology over the other; it is used to identify selected trees, including tree species, measurement of diameter and tree height. Thus, the practical result is tree position, tree species and tree height diameter. Practically, however, it is compared with only one other technology, what I consider to be little insufficient.

In chapter Materials and Methods are describing the species mixed English and Latin names – it should be consolidated (in brackets always Latin names, in cursive). It is not enough clear, what the design should be selected with 3 research plots. Actual description that evokes only a collection with 3x200 trees – there were independent stands with own growth area?  What dimensions (shape, width, length etc) of the plots was used? Fig. 2 left serve to localization of research area and is suitable, in right, however, does not explain at all.  I propose to add in right design of three research plots in the picture. Fig. 3 has maybe a relatively low resolution or readability for the printed version, I suggest it be edited or omitted (consider the ability to notice and use in this work). Instead the term "standing wood" you can use term "standing trees". Fig. 5 it is also difficult to read for the printed version. Fig. 6 I do not consider, in the context of the whole text, should be especially necessary.

The results of the comparative measurement of the tree diameter and heights are very promising; the identification of the species of trees has been excellent. Important are the mentions of disadvantages in presented technology and its limitations for certain stand types or individual trees.

In chapter Results (3.4) and table 4 I recommend changing the currency (I do not know what is “RMB”), preferably convert in US dollars or in any other convertible currency.

I recommend to check all References by editor and Journal requirements, when in more than one absence such as number of pages. Unclear is source number 17, the sources 34 and 37 are the same.

Author Response

Point 1: Contribution of same authors was on base of reviewers comments was rejected, significantly changed and modified. In spite of my statements in the previous review, now I have suggestion to the level of English. The abstract would need to be revised (revised from native speaker).  I think it is no good three times repeat… are used to obtain ... "in the first sentence. Introduction was modified and abbreviations were explained, it is OK. Instead term “altimeter” I would use hypsometer or height meter. It was not measure altitude, but the tree height. Purposes of your work would I would have once modified or specify, the goal is to present a new cost-saving technology over the other; it is used to identify selected trees, including tree species, measurement of diameter and tree height. Thus, the practical result is tree position, tree species and tree height diameter. Practically, however, it is compared with only one other technology, what I consider to be little insufficient.

Response 1: The re-submitted manuscript substantially modifies all chapters of the article, and describes the research work of the paper in more accurate language, using "hypsometer" instead of "altimeter".

The purpose of this study is to propose a new cost-saving technology for forest GIS survey. The background of this technology is the scene in which field workers use mobile phones to collect forest GIS data.

The research background of this paper is derived from the case of “mobile GIS application in forest survey”. Mobile GIS is a GIS system supported by the mobile Internet, with mobile phones or tablet computers as terminals, combined with beidou, GPS or base station as positioning means. Field workers in the forest use mobile phones equipped with GIS system for data collection and GNSS function positioning of mobile phones to obtain geographic coordinates in real time. When using a smartphone to collect data, you need to draw the position of the tree on the map of the mobile phone. The present is a point on the map (geometry). We also need to enter the attribute data for the drawn point (tree position),such as DBH, tree height, coordinates, tree species, small class number, tree number, canopy begins, canopy height, canopy diameter, perimeter breast height, tree health and so on a series of data. These attribute data need to be collected by field workers , among which measuring DBH and tree height is the most time-consuming and laborious. Field workers usually uses hypsometer or total stations to measure tree height, and calipers or diameter tapes to measure DBH. Finally, these attribute data are input into the mobile phone's GIS software.

  With regard to the research on the collection of forest resource data by mobile GIS, in order to let the distinguished experts better understand the operation process, I have provided the reference content of the following link.

。

https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/II-2-W2/165/2015/isprsannals-II-2-W2-165-2015.pdf

https://www.sciencedirect.com/science/article/pii/S221201731300073X

http://www.supergeotek.com/index.php/201602_forestapp/

https://link.springer.com/article/10.1007/s11676-011-0160-1

https://www.fs.fed.us/news/releases/us-forest-service-offers-new-digital-maps-mobile-devices

https://nfg.no/wp-content/uploads/2019/01/ForestInventory.pdf

https://www.trimble.com/Forestry/Mobile-GIS.aspx

Point 2:

In chapter Materials and Methods are describing the species mixed English and Latin names – it should be consolidated (in brackets always Latin names, in cursive). It is not enough clear, what the design should be selected with 3 research plots. Actual description that evokes only a collection with 3x200 trees – there were independent stands with own growth area?  What dimensions (shape, width, length etc) of the plots was used? Fig. 2 left serve to localization of research area and is suitable, in right, however, does not explain at all.  I propose to add in right design of three research plots in the picture. Fig. 3 has maybe a relatively low resolution or readability for the printed version, I suggest it be edited or omitted (consider the ability to notice and use in this work). Instead the term "standing wood" you can use term "standing trees". Fig. 5 it is also difficult to read for the printed version. Fig. 6 I do not consider, in the context of the whole text, should be especially necessary.

Response 2: The new manuscript is supplemented with the English name and Latin name of the species. The shape and position of the three plots are redrawn on the picture and “standing trees”are used instead of “standing wood”.

Point 3:

In chapter Results (3.4) and table 4 I recommend changing the currency (I do not know what is “RMB”), preferably convert in US dollars or in any other convertible currency.

Response 3: This is the currency of the People's Republic of China, abbreviated as RMB or CNY. For the convenience of reading, this article is uniformly expressed in US dollars.

Point 4:

I recommend to check all References by editor and Journal requirements, when in more than one absence such as number of pages. Unclear is source number 17, the sources 34 and 37 are the same.

Response 4:References were re-inserted and the same references were deleted.

Author Response File: Author Response.docx

Round  2

Reviewer 1 Report

I do appreciate the 'research work' and acknowledge the effort made to reply to the criticisms of previous revisions.
Now I start  to grasp the ideas behind the sentences, still some secrets remain unveiled.

The software is a relevant component of the instrument and it appears to be unavailable/not reproducible. As a consequence the whole work is not reproducible.

The paragraph on the testing of the instrument is not getting to the point at all: data have strange

similar biases!

Best regards

Comments for author File: Comments.pdf

Author Response

Point 1: I do appreciate the 'research work' and acknowledge the effort made to reply to the criticisms of previous revisions.
Now I start to grasp the ideas behind the sentences, still some secrets remain unveiled.

The software is a relevant component of the instrument and it appears to be unavailable/not reproducible. As a consequence the whole work is not reproducible.

Response: The core algorithm and operation flow of the software have been announced in the new manuscript submitted, and the specific configuration of the hardware is also published in "Appendix A". Readers can repeat the research work in this paper.

The functions of the software are achievable and repetitive, and the key algorithms and technical details of the implementation have been described in this paper. In order to let the respected experts have a more detailed understanding of the implementation process of the software, the following contents can be referred to, and the core code of the software algorithm in this paper can be put in a folder and provided to the experts in the form of links. In addition, in order to experts more clear about the software operation process, we specially recorded the video of the software operation for the experts, including the video of the tree species "Demonstrate the function of tree species identification.mp4" and the video measuring the DBH and the tree height "Demonstrate the function of Measuring DBH and tree height .mp4" and placing two videos in the file named "Software development kits and materials".

I used Google Drive to upload relevant materials of the software, which can be accessed and downloaded by respected experts through the following links.

https://drive.google.com/open?id=1ci6ykSbygb9i2-pGOrdWpsl5tMCPhk7T

（1）The function of tree species identification is realized mainly through two steps:

1Transfer Learning and model building on PC

The training parameters and steps can be referred to the "2.5.1" and "2.5.2" sections of this paper. The transfer Learning of tree species recognition model is carried out in the TensorFlow framework on the PC platform. The MobileNets network structure is built using the Python programming language, and the parameters that MobileNets pre-trains on the ImageNet dataset are imported. The experiment was performed on a Core I9 series CPU, 32G RAM, NVIDIA GeForce GTX Titan Z 12G GPU, Linux operating system.

(work station for tree species recognition model migration learning and training)

  (Readers can run the software in this article on a smartphone based on the Android operating system.)

2Deploy the trained model on the mobile phone based on Android operating system

Under the Windows 10 operating system, the application development environment based on Android SDK (Java Development Kit) + Java JDK 8 (Java Development Kit) + Android studio 3.3+ ADT (Android Development Tools) is built. The application is suitable for Android6. 0 and above. The most important key to realize the tree species recognition function is to call the development package and dynamic library of TensorflowLite in the development environment: “libtensorflow_inference.so” and “libandroid_tensorflow_inference_java.jar”, and load the trained tree species recognition model through the functions and interfaces inside. I put these two files in the link above, in the file named "Tree species identification model development kit.rar".

Enter the main interface of tree species identification function.

(Pictures can be taken in the field or selected from the phone album and loaded into the main interface.)

(Click the "RECOGNITION" button to quickly and automatically identify the tree species.)

(2) Achieve the function of measuring DBH and tree height

The algorithm for measuring the DBH and tree height can refer to the “2.4.1” section of this paper.

Under the Windows 10 operating system, the application development environment based on Android SDK (Java Development Kit) + Java JDK 8 (Java Development Kit) + Android studio 3.3+ ADT (Android Development Tools) is built. The application is suitable for Android6. 0 and above. Attach the key code for the DBH and tree height measurement algorithm in the link above, in the file named "Key Codes of the Algorithms for Measuring DBH and Tree Height.rar".

(Load the picture taken into the measurement interface, and move the red "+" shaped line to the laser point, then click the button.)

Move the red "+" shaped line to the left side of the DBH, and click the button. Then move the red "+" shaped line to the right side of the breast diameter, and click the button. The DBH value would appear in the interface.

Move the red "+" shaped line to the top of the tree and click the button. Move the red "+" shaped line to the bottom of the tree and click the button.

The values of DBH and tree height would be displayed in the interface respectively.

Point 2:The paragraph on the testing of the instrument is not getting to the point at all: data have strange similar biases!

Response: After implementing the hardware and software functions, this article has carried out a lot of testing on the instrument. This study was carried out in the artificial forest of the Haidian District (40°00′35″ N, 116°20′10″ E) of Beijing, China.The data obtained by the ground measurement tools were compared and analyzed with those obtained by diameter tape and total stations respectively.

The DBH and heights of 600 trees were measured using a diameter tape and total station, respectively, and the data were taken as the reference values of the DBH and tree height for the ground measurement tool in this paper. Twelve tree species were tested using the trained tree species recognition model, and the results of the artificial identification were used as reference values. Position accuracy, DBH and tree height were evaluated by utilizing the bias, root mean squared error (RMSE), relative bias (rBias), and relative RMSE (rRMSE), which are defined by Equations (1)–(4), respectively.

Bias=        (1)

RMSE=           (2)

rBias%=×100%          (3)

rRMSE%=×100%        (4)

The confusion matrix and accuracy (ACC) were used to verify the accuracy of the tree species recognition model, as shown in Table 1 and Equation 5.

The accuracy calculation equation is
ACC= (16)

Experiments show that the position coordinates of trees have positioning accuracies of (RMSE) 0.135–0.183 m, the relative root mean square error (rRMSE) of DBH was 3.61–4.17%, and when considering forest canopy occlusions, the estimated tree height measurement rRMSE was 3.87–4.74%, which conforms to the traditional forest investigation requirements. The accuracy of the tree species recognition model was 94.02% on a training dataset and 93.21% on a test dataset in the mobile phone.

For other responses, please see the attachment.

Author Response File: Author Response.pdf