A Study on Modular Smart Plant Factory Using Morphological Image Processing

Round 1

Reviewer 1 Report

1. The background and current state of the research is analyzed less in Introduction and fails to analyze how this paper differs from other papers.
2. Most of the diagrams in the paper are not very clear and replacement is recommended.
3. “If the photovoltaic power generation performance is insufficient, the motor is driven to fold and fold the solar modules”, what will lead to insufficient photovoltaic power generation performance? And what are the effects of folding solar modules and what are the benefits? Proposed addition of documentation support.
4. The content of the paper involves little image processing and does not use data to show the results of processing, it is not appropriate to use morphological image processing in the title.
5. “Plant systems grow plants using LED light as an artificial light source instead of the sunlight needed for photosynthesis”, there is no comparative experiment to prove that artificial light is better than sunlight. Proposed addition of documentation support.
6. Section “Image processing of plant factory”, content of specific algorithms is missing.
7. “lettuce was separated from the seedling plate by applying proposed morphological image processing”, What kind of morphological image processing is proposed that is not seen in the paper?
8. “Information about lettuce can be obtained by calculating the area of the image”, what is the specific experimental procedure for calculating the area to obtain plant information?
9. “As a result, it was possible to remove some of the stony leaves in the filtered image of Figure 16”，the process of removing the stony leaves is too brief,
10. “First, the intelligent solar module was to adjust the angle according to the situation, as a result, it can be seen that the energy efficiency is improved”, but the process of adjusting the Angle does not have a comparative experiment of different angles.
11. Line 262 should be Table 4.
12. “The LED lamp was made to be 36.7 as a result of PPFD”, what does “36.7” mean?

Author Response

First of all, I would like to thank you for reviewing insufficient thesis carefully and pointing out the wrong part. The parts pointed out by your judges were revised and supplemented as follows and reflected 100% in the original thesis. Thank you.

1. The background and current state of the research is analyzed less in Introduction and fails to analyze how this paper differs from other papers.

- We added content to the introduction, and the reference is number 25.

2. Most of the diagrams in the paper are not very clear and replacement is recommended.

- Fig. 1, Fig. 2, Fig. 4, Table 1, Fig. 7, Fig. 11, Fig. 12, Fig. 13 were changed and reflected in the thesis.

3. “If the photovoltaic power generation performance is insufficient, the motor is driven to fold and fold the solar modules”, what will lead to insufficient photovoltaic power generation performance? And what are the effects of folding solar modules and what are the benefits? Proposed addition of documentation support.

- Line 82 has been added by reflecting the contents of the point of interest.

4. The content of the paper involves little image processing and does not use data to show the results of processing, it is not appropriate to use morphological image processing in the title.

- In this paper, the description of image processing is written from page 8 to page 10. The main contents were to propose a morphological algorithm when separating lettuce seedlings and to use it to determine the shipping time by calculating the area.

5. “Plant systems grow plants using LED light as an artificial light source instead of the sunlight needed for photosynthesis”, there is no comparative experiment to prove that artificial light is better than sunlight. Proposed addition of documentation support.

- Although sunlight is used for plant growth only between sunrise and sunset, LED light sources can perform photosynthesis of plants by supplying electricity for 24 hours. And, it can be used in underground spaces where natural light does not enter.

6. Section “Image processing of plant factory”, content of specific algorithms is missing.

-Created by adding a morphological algorithm to line 208.

7. “lettuce was separated from the seedling plate by applying proposed morphological image processing”, What kind of morphological image processing is proposed that is not seen in the paper?

- Figures 16 and 17 of this paper are presented, and filtering and morphological algorithms are applied as a method.

8. “Information about lettuce can be obtained by calculating the area of the image”, what is the specific experimental procedure for calculating the area to obtain plant information?

-Figure 21 shows the segmentation for lettuce seedlings, and the area is calculated by the size of pixels by extracting only the green part.

9. “As a result, it was possible to remove some of the stony leaves in the filtered image of Figure 16”，the process of removing the stony leaves is too brief,

- Fig. 16 shows the image segmentation process and filtering is performed first. And by applying the added morphological algorithm, you can get an image like Figure 17.

10“First, the intelligent solar module was to adjust the angle according to the situation, as a result, it can be seen that the energy efficiency is improved”, but the process of adjusting the Angle does not have a comparative experiment of different angles.

- When writing the paper, the experimental results for the 100 degree angle of the solar module that were missing are added to Table 2.

11 Line 262 should be Table 4.

- It has been modified and reflected in table 3 --> table 4.

12 “The LED lamp was made to be 36.7 as a result of PPFD”, what does “36.7” mean?

-PPFD (Photosynthetic Photon Flux Density) refers to the number of squares falling per unit time per unit area. The PPFD of the LED light source produced in the text has an effect on the growth of plants and is designed according to 36.7 for rapid growth of lettuce seedlings.

Reviewer 2 Report

This paper presents a practical application of image processing methods for smart plant factory. 

1 seems like the background is important to the morphological processing, might need to elaborate on robustness issues, e.g. mixed or complex background.

2. Some concerns about the cost efficiency. Doesn't look like it could be easily deployed at the larger scale. Please give some more explanation.

3. Figures are not clear enough, please try to use vector format if possible. e.g. the figure in Table 1. is very blurry, hard to tell x, y axis.

4. Lacks an overall flowchart of the whole system, so please can get the big picture easily.

 

Author Response

First of all, I would like to thank you for reviewing insufficient thesis carefully and pointing out the wrong part. The parts pointed out by your judges were revised and supplemented as follows and reflected 100% in the original thesis. Thank you.

1 seems like the background is important to the morphological processing, might need to elaborate on robustness issues, e.g. mixed or complex background.

-Created by adding a morphological algorithm to line 208.

2. Some concerns about the cost efficiency. Doesn't look like it could be easily deployed at the larger scale. Please give some more explanation.

-The proposed method is currently delivered to high school and used for student practice, and it is proposed to be used in general homes.

3. Figures are not clear enough, please try to use vector format if possible. e.g. the figure in Table 1. is very blurry, hard to tell x, y axis.

- Fig. 1, Fig. 2, Fig. 4, Table 1, Fig. 7, Fig. 11, Fig. 12, Fig. 13 were changed and reflected in the thesis.

4. Lacks an overall flowchart of the whole system, so please can get the big picture easily.

- The whole system is added to Figure 19.

Reviewer 3 Report

The authors have identified a good topic of research and presented a decent study. While the work and results are intriguing, this manuscript has several limitations.

1. The manuscript title stated "...modular smart plant factory ..." However, the concept of modularity or factory are not clear. A prototype was done in a lab of limited scope.
2. The unique or novel contributions in the presented research is not clear. The modular IoT architecture has been published in numerous articles. The image processing algorithms used to study plant yield have extensively published in the past. 
3. Some of the figures are very difficult to read. Example. Figure 1, 7, etc.
4. Source of a few figures are not clear. Were these figures obtained from published literature? If so, please cite them accordingly.
5. Numerous typographical errors exist throughout the manuscript. It could use an extensive proof reading.
6. If the novel contribution is improvement in yield or power consumption, this manuscript needs to compare the results with published literature.
7. Literature review need to be expanded to include recent work. examples:

https://www.mdpi.com/2079-9292/7/7/114

https://www.mdpi.com/2079-9292/9/5/726

https://www.mdpi.com/1424-8220/19/11/2484

 

Author Response

First of all, I would like to thank you for reviewing insufficient thesis carefully and pointing out the wrong part. The parts pointed out by your judges were revised and supplemented as follows and reflected 100% in the original thesis. Thank you.

1.The manuscript title stated "...modular smart plant factory ..." However, the concept of modularity or factory are not clear. A prototype was done in a lab of limited scope.

-The small ones produced in this paper are correct. However, if this system is expanded, it is determined that it can be used as a smart plant factory, so the thesis title was written.

2. The unique or novel contributions in the presented research is not clear. The modular IoT architecture has been published in numerous articles. The image processing algorithms used to study plant yield have extensively published in the past.

-The proposed method improves the performance compared to the previous one by utilizing LED design, folding solar module design, and morphological algorithm.

3. Some of the figures are very difficult to read. Example. Figure 1, 7, etc.

- Fig. 1, Fig. 2, Fig. 4, Table 1, Fig. 7, Fig. 11, Fig. 12, Fig. 13 were changed and reflected in the thesis.

4. Source of a few figures are not clear. Were these figures obtained from published literature? If so, please cite them accordingly.

- Most of the values presented in this paper were obtained through experiments, and the figures in Figure 1 and Table 1 were added to those of reference [6][13].

5. Numerous typographical errors exist throughout the manuscript. It could use an extensive proof reading.

- All errors and typos for this paper were reflected in the proofreading paper.

6. If the novel contribution is improvement in yield or power consumption, this manuscript needs to compare the results with published literature.

- In this paper, intelligent solar panels, LED modules suitable for plant growth, and new plant delivery decisions through image processing, were not reflected in existing references because they could not be found consistent with the contents of this paper.

7.Literature review need to be expanded to include recent work. examples:

- The latest thesis was additionally reflected by reflecting the pointed out. reference [28]

Round 2

Reviewer 2 Report

I can see some improvement after revision. However some of the comments have not been fully addressed.

1) The process of morphological processing has been added, but what I was asking is actually about the robustness of morph processing for this application, suppose there are a lot of background clusters or the foreground and background has less intensity difference.

2) The flowchart in Fig. 19 has been added to show the big picture, but there is no caption with detailed explanation.

3) Please add some more reference on using deep learning methods for image-based segmentation, since it's very popular now. Maybe cite the following papers and have some discussion about how it might help your future work, in regards of segmentation.

https://link.springer.com/article/10.1007/s11042-020-09691-y

https://ieeexplore.ieee.org/document/7298965

4) The cost efficiency has not been explained. Looks like the system it over complicated on the hardware side, with an implementation of simple image-based algorithms. 

 

Author Response

The completion of our thesis seems to have improved due to the faithful review of the judges. The points pointed out by the judges were revised and supplemented as follows and reflected 100% in the original paper. Thank you.

1) The process of morphological processing has been added, but what I was asking is actually about the robustness of morph processing for this application, suppose there are a lot of background clusters or the foreground and background has less intensity difference.

-In the lettuce seedlings applied in this paper, there was a difference between the object and the background, so it was possible to separate the object and the background by applying a simple morphological algorithm. In future research, we plan to apply domain division that combines the development of improved morphological algorithms and domain growth methods by reflecting the opinions of the judges.

 

2) The flowchart in Fig. 19 has been added to show the big picture, but there is no caption with detailed explanation.

-[248 line] A detailed description of Figure 19 has been added to the text.

 

3) Please add some more reference on using deep learning methods for image-based segmentation, since it's very popular now. Maybe cite the following papers and have some discussion about how it might help your future work, in regards of segmentation.

https://link.springer.com/article/10.1007/s11042-020-09691-y

https://ieeexplore.ieee.org/document/7298965

 

-References recommended by the judges were additionally reflected in this paper. We will continue to conduct research to upgrade the image segmentation by checking and verifying the recommended references. [Reference – 30,31]

 

4) The cost efficiency has not been explained. Looks like the system it over complicated on the hardware side, with an implementation of simple image-based algorithms.

- [322 line] Details related to costs were written and reflected in the conclusion of the main text.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

I do notice that some improvements have been made to the manuscript, but this revision has not sufficiently addressed all the concerns stated.

1. The author states that using LED lights, folding solar module, and morphological algorithms is a novel contribution. First, using LED lights and folding solar module are not novel, and have been published extensively. The morphological algorithm might be novel, but its respective efficiency in comparison to existing literature is not clear.
2. Some of the figures are very hard to read (e.g Figure 2 and Figure 3)
3. The literature review is not comprehensive. A few suggestions were made, but were not incorporated into the manuscript.
4. Several typos still exist, and this manuscript requires extensive proofreading. Example: “Hard Ware” in Figure 9 should be “Hardware”
5. Can you change the background color of Fig. 20. It is difficult to read the waveform.

Author Response

The completion of our thesis seems to have improved due to the faithful review of the judges. The points pointed out by the judges were revised and supplemented as follows and reflected 100% in the original paper. Thank you.

1.The author states that using LED lights, folding solar module, and morphological algorithms is a novel contribution. First, using LED lights and folding solar module are not novel, and have been published extensively. The morphological algorithm might be novel, but its respective efficiency in comparison to existing literature is not clear.

-The proposed thesis is that the modular smart plant factory system using intelligent solar modules has higher efficiency compared to the existing system by improving energy efficiency, rapid growth of plants, and determining the shipping timing through image processing. Therefore, it can be said that the proposed system is improved more efficiently than the existing system.

 

2. Some of the figures are very hard to read (e.g Figure 2 and Figure 3)

-The clarity of Figure 2 and Figure 3 has been increased and reflected in the thesis.

 

3. The literature review is not comprehensive. A few suggestions were made, but were not incorporated into the manuscript.

      -References[29] related to this paper have been added.

 

4. Several typos still exist, and this manuscript requires extensive proofreading. Example: “Hard Ware” in Figure 9 should be “Hardware”

      -The errors in Figure 9 have been corrected and reflected in the thesis.

 

5. Can you change the background color of Fig. 20. It is difficult to read the waveform.

     -The sharpness of Fig. 20 was increased and reflected in the thesis.

Author Response File: Author Response.pdf