Deep Learning-Based Occlusion Handling of Overlapped Plants for Robotic Grasping
Round 1
Reviewer 1 Report
In this paper, a novel method is presented using segmentation-based handling of occluded plants on flat surfaces such as conveyor belts to determine the optimal grasp. This method enables the robotic grasping of plants in an application that involves autonomous manipulation of plants, such as robotic plant propagation. The method consists of using a CNN for handling plant occlusions and a method for detecting grasps based on the segmentation results of the CNN.
I consider the presented article to be original and it contains an adequate amount of used literature. The conclusion is relatively short. This section does not reflect the entire contents of the manuscript, including the analysis of the results obtained. This section should be extended. The scientific benefits of this article should be explicitly mentioned in the article. The present article does not state in which authors they plan to continue the research. It is clear from the presented paper that the research will need to be continued.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 2 Report
The article emphasis on the problem of occlusion in the plant species. Author have detailed the problem statement related to vision-based segmentation which leads to the problem of occlusion, but if authors include a set of research questions identified for this study, then there will be clarity in proposed research methods addressing the problems.
The motivation is not clear. Please specify the importance of the proposed solution in the introduction section.
Related works looks like a summary of existing works, author must revise the section by including the strength and limitation of each work with a achieved results. Include a section on how the proposed method overcome the gaps in the existing works.
Formalize the grasp detection algorithm and include the overall algorithm design.
Compare the proposed solution with more recently published work
Discuss the limitations of the proposed method. Every method has a weakness and should be discussed clearly and give solutions in the future work section
Author Response
Please see the attachment
Author Response File: Author Response.docx