Multidimensional Spatial Vitality Automated Monitoring Method for Public Open Spaces Based on Computer Vision Technology: Case Study of Nanjing’s Daxing Palace Square

Comments 1: Authors explain research procedure on the site, but don't explain if the selection of site was important and what were criteria of the selection.

Response 1: Thank you for pointing this out. The selection of the research site is crucial, as it determines whether our study can test the performance of spatial vitality detection methods across a wider range of space and activity types, and conduct method exploration at lower experimental costs. In the preliminary design phase of our research, we first clarified that the research objective was to explore the feasibility of spatial vitality measurement methods. Based on this, we established criteria for selecting the research site.

Firstly, since method exploration requires repeated experimental trials, the research site should be of appropriate size to ensure that utilizing multiple camera positions to cover the research area will not incur excessively high research costs.

Secondly, the research site should have a high level of activity richness to validate the stability of the spatial vitality monitoring method when analysing different trajectory features.

Thirdly, the research site should contain spaces with varying spatial features, which aids in testing the stability of the spatial vitality monitoring method under various spatial characteristics and in exploring the differences in spatial vitality across different spatial features.

Therefore, based on the characteristics of Daxing Palace Square (See lines 138-146 for details), the research ultimately chose it as our research site.

In Section 2.2, we briefly introduced the reasons for selecting Daxing Palace Urban Square as the research site. In the revision, we added further explanations for the criteria for selecting the research site. Additionally, in the discussion on research limitations in section 4, lines 562-576 include a description of the limitations of the research site.

Comments 2: It is not clear why in the staying behaviour identyfication process (Fig. 7) there is non-staying point represented.

Response 2: Thanks for the useful suggestion. In the stay point identification algorithm designed in this paper, we assign a label to each trajectory feature point indicating whether it is a stay point, facilitating subsequent calculations. The "Staying Point" and "Non-staying Point" process modules shown in Figure. 7 can be regarded as the two outcomes of trajectory points after undergoing stay point identification, categorized as either "Staying Point" or "Non-staying Point". The arrow from the "Non-staying Point" process module pointing to the first module signifies the selection of the next new trajectory feature point following one that has been identified as a "Non-staying Point" for subsequent identification, denoted by the process "m=m+1". This does not imply that the "Non-staying Point" is added to the stay point identification process. In the revised version, we have improved this image based on the identification code logic designed in this paper to display the identification process more clearly.

Comments 3: Page 9: Reviewer doubt whether it is right thing to threat point as feature of movement trajectory: trajectory is a dynamic fenomena, point is static one.

Response 3: In response to this question, we have revised the discussion in lines 251-253 to eliminate any ambiguity arising from the wording. Specifically, we have recorded the characteristics of the trajectory in the form of trajectory feature points with multi-dimensional attributes. Therefore, the trajectory feature points serve as carriers for recording the dynamic characteristics of the trajectory across four dimensions of vitality, rather than being the trajectory features themselves. We’ve changed “Combined with staying behavior point data, the trajectory points of staying behavior reflect the intensity of human activity, whereas those of non-staying behavior reflect people's transit speed.” to “When integrated with the results of staying behavior recognition, the movement speed of staying behaviors indicates the intensity of human activities, while the movement speed data for non-staying behaviors represents the transit speed of individuals.”

Comments 4: Page 10: what is the exact the difference between structural differences and significant structural differences?

Response 4: This was a mistake in the wording. We did not intend to emphasize the diffidence in the level of significance between two “structural differences”, but rather to illustrate that while structural differences exist between different trajectories, the structure within a single trajectory is not homogeneous across its different parts. In the revised manuscript, lines 276-278, we have corrected this error and conveyed the correct meaning. We’ve changed “Regarding the quantification of the behavioral structures, structural differences exist for different behavioral trajectories, while significant structural differences exist among the local behavioral trajectory segments at different periods.” to “Regarding the quantification of the trajectory structures, structural differences exist for different trajectories, while structural differences exist among the local structures within the same trajectory.”

Additionally, this paper utilizes trajectory structure to calculate the complexity of behavior, thereby characterizing Spatial Participation (see Section 2.4.4). Therefore, in the entire document, we have replaced the ambiguous term "behavioral structure" with "trajectory structure." Simultaneously, we have unified all instances of "trajectory point" and "feature point" to "trajectory feature point" to avoid conceptual confusion. Please review and check it.

Comments 5: In some aspects the research don't use universal indicators: for example in some cultures nobody dance in the public space.

Response 5: Based on our understanding, the reviewer might have had concerns about the statements in lines 273, 463, 482, and similar sentences. In the statement in line 273, our wording may have caused ambiguity. Dancing and any specific type of activity were not used as indicators for quantifying spatial vitality. What we intended to convey is that people's deep engagement with a space enhances its vitality. We measure spatial participation through the calculation of trajectory features and consider it a factor of spatial vitality. Activities like dancing, which involve deep participation with space, thus serve as examples of high spatial participation behaviors. To clarify this point, We have revised the original text in lines 269-273 from “Regarding crowd behavior, using space as a conduit for transportation results in a simple trajectory structure, whereas complexity implies that behavior is influenced by a greater number of environmental or societal factors. This suggests that people's deep participation in space, such as resting, enjoying the scenery, or dancing, enhances their spatial vitality.” to “Regarding crowd behavior, using space as a conduit for transportation results in a simple trajectory structure. In contrast, behaviors such as sitting, sightseeing, dancing, and others exhibit complex trajectory structures, indicating that these behaviors are subject to a greater extent to environmental or social forces. This suggests that people's deep participation in space, thereby enhancing spatial vitality.”

In lines 463 and 482, square dancing similarly appears as an example rather than an indicator. To eliminate ambiguity, we have revised the original text. In line 463, we changed “Influenced by aggregative crowds, represented by square dancing, the staying behavior ratio value in s3 at 08:30 is second only to that of s6.” to “Influenced by aggregative crowds, such as square dancing, the staying behavior ratio value in s3 at 08:30 is second only to that of s6.” In line 482, we changed “This suggests that low-density, low-speed behaviors tend to occur in high-enclosure spaces with better rest facilities, whereas high-density, low-speed behaviors, represented by square dancing, tend to occur in low-enclosure spaces.” To “This suggests that low-density, low-speed behaviors tend to occur in high-enclosure spaces with better rest facilities, whereas high-density, low-speed behaviors, such as square dancing, tend to occur in low-enclosure spaces.”

Moreover, we must acknowledge that the conclusions drawn from the relationship between spatial features and spatial vitality observed in this study's site may not be applicable to all cultural backgrounds and regions. However, the quantified results of spatial vitality from this study are generalizable. In Section 4, we have added a more detailed discussion on this issue for the reviewer's consideration.

Comments 6: Lines 543/544: there probably is a mistake: sentence in line 543 finishes with a drop, but text in next line (544) isn't initialed by a capital letter.

Response 6: Thank you for pointing this out. The sentence is currently located in lines 551-552. The period was a writing error, we’ve changed “Regarding the coupling between spatial vitality and spatial features, our experimental results had strong explanatory power for the "boundary effect" theory, indicating a significantly positive impact of landscape boundaries on spatial vitality. phenomenon through our statistical analysis of the spatial vitality factor calculations and spatial vitality distribution maps supported by discrete feature point data.” to “Regarding the correlation between spatial vitality and spatial features, our experimental results had strong explanatory power for the "boundary effect" theory, indicating a significantly positive impact of landscape boundaries on spatial vitality phenomenon. This was deduced through our statistical analysis of the spatial vitality factor calculations and spatial vitality distribution maps supported by discrete trajectory feature point data.” We corrected the incorrect use of periods and broke down long sentences to enhance readability. Additionally, we’ve changed " Regarding the coupling between spatial vitality and spatial features," to " Regarding the correlation between spatial vitality and spatial features," to reflect the intended meaning more accurately.

Comments 1: Line 35: The citation format for Jacob (1961) needs revision to align with the standard citation guidelines.

Response 1: Thank you for pointing this out. In line 35, we have already made the necessary corrections to the citation format. Similarly, we corrected the same error in line 65. Please review and check it.

Comments 2: Line 108: There is a need to define the software and equipment used in the study more clearly.

Response 2: Agree. In lines 110-114, we have added supplementary explanations: “Specifically, we deployed the YOLOv7-DeepSort model on the Python platform to extract pedestrian trajectories from videos. We then developed a method for quantifying spatial vitality from four dimensions: crowd heat, staying behavior ratio, movement speed, and spatial participation. For detailed methodologies, see Section 2.3 and 2.4.” Furthermore, Sections 2.3 and 2.4 provide detailed descriptions of the data collection devices used in the study and the computer software employed for video trajectory extraction. Please review and check it.

Comments 3: Figures 2, 10, 13, 16, and 19: These figures require text readability and scale improvements. The current presentation is not clear enough.

Response 3: Thanks for the useful suggestion. Regarding Figure 2, we have changed the font and size to improve the readability of the text. Additionally, we have altered the presentation of the satellite image to emphasize the research site more clearly. For Figures 10, 13, 16, and 19, we have enlarged the images in the layout and labeled the spatial areas. At the same time, we have adjusted the color scheme and font in Figures 1, 3, 4, 5, and 6 to highlight the main content of the images.

Comments 4: Study Limitations: A more detailed explanation of the study's limitations is necessary, along with a discussion of variables that should be considered in different contexts.

Response 4: We have revised the discussion section of the paper. In lines 539-548, we discuss the potential impacts of data loss due to trajectory tracking loss, highlighting the requirement of this spatial vitality measurement method for video quality. In lines 562-602, we summarize the limitations of this study. Firstly, we acknowledge the limitations in the generalizability of conclusions due to the singularity of the research site and provide examples of variables that might need to be considered in different types of public open spaces. Secondly, we outlined potential directions for future research on the correlation between spatial features and spatial vitality, as well as spatial features that may need to be considered when applying the methods of this study to other types of public open spaces. Lastly, we discuss the causes of sampling errors and possible solutions.

At the end of the conclusion, in lines 641-647 we have added a report on the potential applications of the method developed in this study in the field of urban design and renovation.