A Rapid Recognition Method for Electronic Components Based on the Improved YOLO-V3 Network

Round 1

Reviewer 1 Report

The manuscript is in general well written and scientifically sounds. The problem analysed by the Authors is interesting and important. The provided introduction, state-of-the-art analysis, description of the proposed solution as well as the obtained experimental results are described properly and adequately.

I have only some minor remarks:

- on page 4th and 5th the Authors describe the algorithms for the augmentation. The specific parameters of applied methods are required. Only for brightness transformation something is given.

- some small corrections are needed. The Authors are advised to carefully analyse the final version of the submission. An example on p. 5, 52: "components are not be considered". Several other problems, including grammatical errors can be found in the text.

- The Figure 5 does not give any additional or important information. It is unnecessary.

Author Response

The author's answer to review opinions of the manuscript ID: electronics-546842

I am very grateful to your comments for the manuscript. According to your advice, I amended the relevant part in manuscript. The questions were answered below.

Point 1: on page 4th and 5th the Authors describe the algorithms for the augmentation. The specific parameters of applied methods are required. Only for brightness transformation something is given. 

Response 1: I check specific parameters of three other algorithms and add them to the paper. Please refer to line 137,138,144,145,156,157 on page 4,5. 

Point 2: some small corrections are needed. The Authors are advised to carefully analyse the final version of the submission. An example on p. 5, 52: "components are not be considered". Several other problems, including grammatical errors can be found in the text. 

Response 2: Thanks a lot for the suggestions. The final version of the submission has been analysed again and some small mistakes have been corrected. I invite a native English-speaking people to help me check the grammatical errors and add his name to authors. 

Point 3: The Figure 5 does not give any additional or important information. It is unnecessary. Response 3: Figure 5 has been removed. Please refer to line 175 on page 5.

Reviewer 2 Report

This paper deals with the automatic recognition of electronic components in order to develop intelligent manufacturing systems. 

200 target images were collected from four different electronic components. 

four data augmentation technologies were used: contrast enhancement processing, add 108 noise processing, brightness transformation and blur processing.

A combination of YOLO-V3 157 and Mobilenet was exploited to accomplish the recognition task. Results are compared with some different models and the analysis of the impacts of each step in the whole system are reported. 

I liked the paper, it is very well written and each step is well detailed. 

I would like just to suggest to add a reference to a recent survey on deep learning for assistive applications (like that described in the paper) : 

Deep Learning for Assistive Computer Vision. In Proceedings of the European Conference on Computer Vision  ECCV2018.

Author Response

The author's answer to review opinions of the manuscript ID: electronics-546842

I am very grateful to your comments for the manuscript. According to your advice, I amended the relevant part in manuscript. The questions were answered below.

Point 1: I would like just to suggest to add a reference to a recent survey on deep learning for assistive applications (like that described in the paper):

Deep Learning for Assistive Computer Vision. In Proceedings of the European Conference on Computer Vision  ECCV2018.


Response 1: Thanks a lot for the suggestion. I read this paper carefully and find this paper to be more consistent with the meaning of the statement in line 63 compared to reference [7]. So, I decide to use this paper to replace reference [7]. Please refer to line 63,486.