Deep and Machine Learning Image Classification of Coastal Wetlands Using Unpiloted Aircraft System Multispectral Images and Lidar Datasets

Round 1

Reviewer 1 Report

The authors evaluated the performance of four machine learning techniques on UAS multispectral aerial imagery and accessible lidar dataset. The author compares the differences of the above four methods on multiple data sets, and provides a stable solution for the image classification task. I suggest accepting it after revision. Detailed suggestions are shown as follows.

1.     In section 2.5, more details about the segment mean shift algorithm should be mentioned. The criterion of manually selecting polygons in the algorithm is also confusing.

2.     The training process of traditional machine learning methods is not very clear, especially how the training data are fused into SVM and RF. Drawing a sub graph may better describe the training process.

3.     Some recent works about landcover classification are suggested to cite, such as “Deep Autoencoder for Hyperspectral Unmixing via Global-Local Smoothing” and “Structure-Color Preserving Network for Hyperspectral Image Super Resolution”

4.     For experiment1, more band combinations should be tested to eliminate the influence of random selection.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors,

the Article is relevant and shows significant  findings. However, there is lack of methodological results in the proper section.

Good luck! 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Dear Authors,

I have reviewed the paper "Deep and machine learning image classification of coastal wetlands using unpiloted aircraft system multispectral images and lidar datasets". The aims of the paper are germane with Remote sensing applications topic, in this form of article fits with the international scientific standards.  The contribution of this paper to the scientific knowledge is good. In my opinion, there are no important corrections to be made, just a couple of comments:

- please try to shorten a bit the abstract being more concise

 

 

- pay attention to typing errors (for instance space before "." or ",")

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

My comments:
1. The topic of this paper is interesting and innovate and it will contribute in related research field.

2. A section of “Related Works” or “Literature Review” is necessary for this paper.

3. “2. Materials and Methods” and “3. Results” are well presented.

4. The “4. Discussion” is too brief and should be more detailed.

5. The “5. Conclusions” must be reinforced more. For example, the contributions to academic research as well as theoretical implications and research limitations.

Author Response

Please see the attachment.

Author Response File: Author Response.docx