Hyperspectral Detection of Moisture Content in Rice Straw Nutrient Bowl Trays Based on PSO-SVR
Round 1
Reviewer 1 Report
Please find the attachment.
Comments for author File: 
Comments.pdf
Moderate editing of English language is needed.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
1. The title of this manuscript is not proper. This paper uses hyperspectral imaging technology, compared RF, PSO-SVR and XGBoost to detect the moisture content of Rice Straw Nutrient Bowl Trays. The research results indicate that the PSO-SVR model has the best predictive performance. The article is actually a comparative study of methods, preferably explicit on the title.
2. The paper uses 204 samples collected for model analysis. The sample is insufficient and the condition is single, and it can be seen from Figure 4 that the samples lack actual interference factors and are not representative. Therefore, it is reasonable that the later results appear with higher accuracy.
3. The article uses hyperspectral imaging equipment for data acquisition. The process of the experiment is not clear, such as how to achieve area imaging with the equipment (It requires a relative movement), and what are the lighting conditions? From the experimental process in this paper, the use of fiber spectrometer can also achieve the goal, and the latter has a higher signal-to-noise ratio, which is more conducive to ensuring the accuracy of prediction.
4. Figure 2 only has an image of ROI selection, and there should be another image without ROI for comparison.
5. Hyperspectral preprocessing tends to filter out valid information, and Figure 5 shows that too much preprocessing is not required.
6. The analysis of the Discussion section is insufficient. For example, how applicable is the model? The band of hyperspectral detection of water content is generally in the short-wave infrared band, the paper uses visible/near-infrared band. The manuscript does not analyzed the characteristics of the band. The results are just statistically consistent, the mechanism is not very clear, it is difficult to apply to other places or other scenes.
7. The ordinate of Figure 4 is that the reflectance value is 0-1000, which needs to be modified, generally between 0-1.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Authors have made significant changes in revised version based on reviewer comments. Revised version quality is improved significantly.
Minor editing of English language required.
Reviewer 2 Report
OK now.
