The Combined Use of UAV-Based RGB and DEM Images for the Detection and Delineation of Orange Tree Crowns with Mask R-CNN: An Approach of Labeling and Unified Framework

Round 1

Reviewer 1 Report

This paper evaluated the use of images captured by UAVs and photogramme try techniques using the algorithm Structure from Motion (SfM) for the identification and delineation of treetops located in different spatial densities using deep learning. This paper is significantly novel and the contributions are good for a journal paper. However, the following corrections to be made before consider this paper for publication.

1. There are several performance metrics in the literature, but the authors are considered only few in the paper. It is recommended to provide more metrics considering different datasets.

2. The limitations and future scope of this work must be discussed.

3. Computational complexity and efficiencies are estimated and compared using the existing works.

4. The references from 2022 must be considered in the literature.

5. Highlight the contributions of the paper in Introduction.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

1. The DEM in the abstract should have the full name because it appears for the first time.

2. Mask R-CNN needs to add some more explanation.

3. Equation (1), what is the unit of CIVE?

4. What is the difference between RGB mode and RGBD mode?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

In this paper, the authors propose an approach combining the exploitation of UAVs and segmentation methods to detect and delineate canopies in orange plantations.

The paper is well written and flows well. The context is indeed interesting. There are few aspects that the authors could address in order to improve the quality of the paper:

- Related work should be improved; in particular, the authors could present an overview of current works and how the proposed approach is positioned with respect to them. Several details should be reported. My suggestion is to improve the part of the Introduction devoted to the related work.
- CNN and deep learning methods in general have been used in several context. I suggest the authors to consider citing the following work that could be improved by means of UAVs: https://doi.org/10.1016/j.compeleceng.2021.107572
- Few typos are present; I suggest the authors to carefully read the paper.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The bugs have been fixed.

Reviewer 3 Report

The authors addressed my concerns.