Information Perception Method for Fruit Trees Based on 2D LiDAR Sensor

Round 1

Reviewer 1 Report

In this paper, the authors propose an algorithm to perceive fruit tree position information by 2D lidar sensor.
The paper has severe inconsistencies in its presentation, and the scientific contribution is not clear.
The introduction section does not provide a good overview of the problem that this work pretends to solve. Also, it is not clear about the benefits of detecting fruit trees within the smart farming scope. The references are not arranged in a logical order, and some of them do not seem relevant to this paper's purpose.
In the Materials and Methods section, the contribution is not clearly stated. For example, if the algorithms have been previously presented in other works, where is the contribution? Only in the integration? Also, there is no information about the implementation and deployment elements.
The results section is not clear, and it does not provide enough elements regarding the performance of the proposed approach.
There is a discussion section, but no conclusion section is a serious flaw for a research paper.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

In this paper, a fruit tree detection method based on density lightning link clustering (LAPO-DBSCAN) was proposed. The student used lightning connection algorithm (LAPO) to obtain the initial clustering center of fruit trees.Then, density-based clustering algorithm (DBSCAN) was used to cluster the data that could not be clustered in the first step. Finally,dynamic DBSCAN is used to detect clustering results and the fruit tree results are saved in the form of coordinates.The algorithm has obvious advantages over the previous algorithms, but there are still some small problems to be improved. 
1. Although a large vocabulary is used to introduce papers related to the framework, the key technical explanations of the framework proposed in this paper do not go far enough. 
2. Maybe introduce some more innovations of this article. 
3. Further details are needed for using the algorithm to fuse with other sensors.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors provide the algorithm to get the position and number of fruit trees. The result shows that the positive detection rate is 96.69 %, the false 18 detection rate is 1.67%, and the average processing time is 1.14 s, which verifies the reliability of the algorithm. However several questions were raised:

1.     Please provide the brand of 2D lidar sensor you used

2.     Please adjust the workflow in Figure 2. Some texts are not precise and too small.

3.     For benchmarking, The CCD Sensor/Camera should be placed nearby 2D Lidar sensor. We don't is it correct or not to detect the object. The author should provide the real figure/object captured using a CCD sensor/Camera to compare the result obtained by the Lidar sensor

4.     There are so many references related to applying the 2D Lidar Sensor and more advanced algorithms not cited in this manuscript. also in this manuscript, the references are not updated.

5.     Major revision is required

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Authors have addressed the comments and concerns from the previous revision.

Still minor findings must be corrected:

- Line 11 (in the abstract) the phrase is not correct "When orchard intelligent equipment navigates in orchards, it is necessary to perceive the surrounding environment."

- Line 115 in Figure 2, the concept of pretreatment is not explained.

- Line 322, please explain in detail the differences between the original data nad the preprocessed data.

- Line 452, what are the shortcomings? 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The author(s)  has already been improved the manuscript, however, several comments not addressed properly (such as the real condition image (JPG); the JPG image should be added along with lidar; provide more evidence).

Why does the author apply the algorithm to the same object (Figures 6, 7, and 8)?

whole  manuscript need to be revised

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf