Detection of Growth Stages of Chilli Plants in a Hydroponic Grower Using Machine Vision and YOLOv8 Deep Learning Algorithms

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presented the applicability of MV technology with DL modelling to detect the growth stages of chilli plants using YOLOv8 networks. The influence of different bird’s eye and side view datasets and different YOLOv8 architectures was analysed. This research work has important research significance and application value.

Overall, the research work was full, the research methods were feasible, and the research conclusions were credible.

Author Response

See attached PDF.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors present the use of Machine Vision (MV) technology combined with Deep Learning (DL) modeling to detect the growth stages of chili plants using YOLOv8 networks. The study investigates the effectiveness of both bird’s eye and side view datasets alongside various YOLOv8 architectures.

To ensure the manuscript reaches its full potential, I would appreciate it if the authors could address the following queries:

1. The manuscript divides the growth stages of chili plants into three classes: Growing, Flowering, and Fruiting. Could the authors elaborate on how these classes were precisely defined and the criteria used to delineate the boundaries between each stage?

2.  The paper states that image augmentation techniques such as mirroring and rotating are employed to expand the training dataset. Could the authors discuss why these specific augmentation methods were preferred over simply increasing the frequency of image collection to obtain more real images? What are the benefits of these synthetic modifications?

3. Were there any preprocessing steps implemented to standardize the images before inputting them into the YOLOv8 models? If so, could you detail these procedures and their impact on model performance?

4. What adjustments would be necessary for the proposed model to be effectively applied to other types of plants or different hydroponic setups? Are there particular characteristics of chili plants that might limit the model's applicability?

5. How does the model ensure accurate training when multiple plants are present in a single image? What techniques are employed to identify and select the correct plant for training purposes?

Author Response

See attached PDF.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

A very interesting and well-documented experiment, almost like a research report.

Although the numerical results show a very high recognition rate, the correctness of the identification cannot be deduced from the pictures attached to the article.

I think that the use of more conclusive images could be more useful for disseminating the obtained results.

Author Response

See attached PDF.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript on the application of machine vision system and YOLOv8 DL algorithms presents an well written study.

More references on similar type of approach need to be included in the introduction to carefully introduce the background to a general audience. In addition, othr reported models need to be compared with this study to validate the model performance.

Author Response

See attached PDF.

Author Response File: Author Response.pdf