Assessing the Homogeneity of Forage Mixtures Using an RGB Camera as Exemplified by Cattle Rations

Round 1

Reviewer 1 Report

This paper assesses the homogeneity of forage mixtures using an RGB camera via using cattle rations. The application scenario is  interesting, but I still have some comments on this paper as follows:

1. As far as I can see, this paper only employs a Matlab image processing toolbox to handle the data. So is there any other technical contributions? A journal paper should not look like a user manual of a toolbox.
2. Some figures are poor plotted such as Fig. 1. Why should a picture of a camera to be some big?
3. Moreover, the figure quality is not so good and please increase the original resolution.
4. The description of the image processing algorithm is missing. Please provide a user implementation guildline.
5. The test software environment should be given in detail, otherwise it may course the failure on some old devices. 
6. There lack some comparison between other image processing methods on the same task.
7. Some recent developed image processing methods should also be reviewed in the introduction section such as:  Active contours driven by region-scalable fitting and optimized laplacian of gaussian energy for
   image segmentation, Signal Processing; Active contours driven by adaptive functions and fuzzy c-means energy for fast image segmentation, Signal Processing; A level set method based on additive bias correction for image segmentation, Expert Systems with Applications

Author Response

Thank you for your review.

We significantly improved the manuscript: 

Proposed image processing methods reviewed in the introduction;

figures were edited.

Other comments were taken into account as well.

Reviewer 2 Report

Interesting application of technologies, in a sort of IoT for agro-food.
The attention to the low cost and sustainability of the process is also of interest.

Author Response

Thank you for your review.

Reviewer 3 Report

Assessing the homogeneity of forage mixtures using an RGB 2 camera as exemplified by cattle rations

This paper aims at developing an alternative algorithm for testing the homogeneity of mixed animal feeds. The paper uses an RGB method from a MATLAB built in app or image analyzer. The method has the potential to be a better option considering it is cheaper and faster than the other methods (including NIRS technology) currently being used for the same purpose.

Comments:

The application is important, and the described workflow appears to be interesting. However, the authors did not do a good job at establishing a standard operating procedure of the method nor were they able to justify the significance of the method as no ground truth or thresholding was provided for the experiment. I have identified the following areas for improvement in the paper:

1. Create a concrete story line in the introduction to establish the importance of the method. The current introduction is random and unorganized, which does not do justice to this effort. For example, lines 112- 133, the authors attempt to state the importance of the new method but it is not clearly stated.
2. Use appropriate graphs and properly describe the content of the graphs and how the result shows any trend to support the argument being made. For example, the graph shown in figure 6 of the paper is not well explained and does not fully justify the assertions.
3. Create a proper thresholding for the experiment as a basis for comparison to show that the method is effective as claimed.
4. Properly detail the methods under the “materials and methods” section to ensure better understanding and reproducibility of the proposed idea. For example, on line 248, the authors are still describing a method under “results”.
5. Conduct additional experiments on more data points or images and not just one or few. There is also a need to show the basis of choosing the number of data points used for authenticity.
6. Clearly show the conclusion and how it ties with the introduction. The current solution does not include any information that suggests that the method works better than the current ones. The only advantage stated is that it “has a lower implementation cost and performs online measurements using IoT technology” from line 365-367.
7. Provide the limitations of the method and possible area of improvement.

Author Response

Thank you for your review.

We significantly improved the manuscript:

figures were edited;

graph description extended;

tied conclusion to introduction and provided the limitations of the method.

Other comments were taken into account as well.

Round 2

Reviewer 1 Report

The authors have addressed all my comments, but do please add high resolution figures.

Reviewer 3 Report

The manuscript was sufficiently improved.