Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript „Dual-channel co-spectroscopy method for non-destructive fruit quality detection“ describes dual-channel co-spectroscopy method to address the high cost of spectrometers in current fruit quality non-destructive detection and sorting equipment.

The title of the manuscript „Dual-channel co-spectroscopy method for non-destructive fruit

quality detection“ is too ambitious and does not reflect the results obtained. The authors studied only Fuji apple fruits and did not study other apple varieties or other types of fruits (e.g., pears, kiwi, apricots, tangerines, etc.), so the title should be changed.

The topic is original and relevant to the field, but the need for such research is questionable. Although the authors justified that it is better to use dual-channel co-spectroscopy and thus reduce the cost of industrial equipment, this is not very convincing. The cost of industrial fruit sorting equipment can be reduced by using technological or structural elements in equipment production. I think using a single-channel spectroscopic instrument in a wide wavelength range (400-1100 nm) is more logical than artificially splitting the wavelength range into two intervals (400-700 nm and 700-1100 nm). The manuscript's authors write in the discussion (lines 305-306), "These results demonstrate that splitting the spectral range using filters for dual-channel co-spectroscopy achieves spectral data quality comparable to the single-channel method.”. So what is the point of splitting the spectra then.

Table 4. PLSR modeling performance for soluble solids content in apples using spectra in the 700– 1100 nm range. However, the table does not present data in the 700-1100 nm spectrum, but in the 400-1100 nm range. Please revise it.

In Table 5 (Spectrometer prices and price changes for different sorting line categories) the prices of different sorting equipment are presented, but these different sorting lines are not characterized by either sorting efficiency or quality indicators. Maybe in a single-line sorting line the fruits are sorted not only by the content of soluble solids content, but also by color indicators? Please revise Table 5.

I believe the conclusions do not correspond to the evidence and arguments presented and do not address the main question posed.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This article is well written (Dual-channel co-spectroscopy method for non-destructive fruit 2 quality detection), but here are some suggestions that will improve the quality of the manuscript

Try to avoid repetition of words both in the title and in the keywords.

In the introduction, before introducing visible and near-infrared spectroscopy technologies, some general information should be given about standard analytical techniques for fruit quality testing, as these are very time-consuming, labor-intensive, and expensive.

Material and methods

It is necessary to give more information about the samples and the number of genotypes or varieties. How did you decide on these? Why Fuji, probably because it comes from a cross between two varieties. Also, explain the stage of development of the fruit you collected, and how you checked that. The developmental stage is the main factor affecting fruit quality.

Line 171 5. Fruit pedicel is the correct term, the term you gave is not correct

Lines 174, 179 Clarify section 1.3.1

Provide the information about the software you used for spectral data collection.

Comments on the Quality of English Language

The English could be improved.

Author Response

Please enter "Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors present a study about a novel dual-channel co-spectroscopy method for non-destructive fruit quality detection, especially focused on optimizing spectral data acquisition to enhance sorting efficiency and reduce costs in fruit inspection systems. The paper is strong, technically sound, and well-structured. The proposed method has clear scientific and practical value, with a well-presented discussion supporting the findings. Only minor refinements would be needed for additional clarity and depth.

The introduction effectively provides background on spectroscopy technologies used for non-destructive fruit quality inspection. Also, it presents the novelty of the non-destructive fruit quality detection method based on dual-channel co-spectroscopy.  

The Materials and Methods section in your document is quite detailed and well-structured. It provides comprehensive information regarding: the description of sample conditions, and spectral measurement setup, including optical fibers, halogen lamps, filters, and spectrometers; A self-constructed dual-channel co-spectroscopy platform for non-destructive detection, Data Collection Process, Soluble Solids Content Measurement, Data Analysis & Modeling (Methods for noise reduction and spectral feature selection; Explanation of the regression model (PLSR) used for prediction. Statistical validation procedures (R², RMSE, cross-validation)).  In the Materials and Methods section the authors may consider:

• Clarifying replication, were multiple batches of apples tested, or was this done in one session?
• Software details, if specific software was used for data processing (e.g., MATLAB, Python), mentioning the version could be useful.
• Statistical analysis explanation, while methods like PLSR are introduced, it might help to elaborate on why it was chosen over other models.

The Results section is logically organized, progressing from single-channel PLSR modeling to dual-channel modeling for different spectral ranges. It includes tables summarizing key results, making it easy to compare different preprocessing methods, feature selection techniques, and model performances. Suggestions:

• The text is highly technical and could benefit from a summary of key findings after each subsection.
• Error distributions could further enhance understanding.

The Discussion effectively builds on the results section, explaining spectral trends, model performance, and the advantages of the dual-channel co-spectroscopy approach. It compares single-channel and dual-channel methods in a structured manner, reinforcing the experimental findings. The section summarizes model performance and discusses feasibility, ensuring a smooth transition from the results. I have some suggestions for this section:

• Could explicitly mention how these findings compare to existing research in more detail. It does mention the increasing demand for fruit quality detection, but further comparison with similar studies could add depth.
• A brief mention of limitations (e.g., potential challenges in real-world implementation, and external factors affecting spectral measurements) could provide a balanced perspective.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf