2022
DOI: 10.3390/electronics11193189
|View full text |Cite
|
Sign up to set email alerts
|

Research on Surface Defect Detection of Camera Module Lens Based on YOLOv5s-Small-Target

Abstract: For the problem of low resolution of camera module lens surface defect image, small target and blurred defect details leading to low detection accuracy, a camera module lens surface defect detection algorithm YOLOv5s-Defect based on improved YOLOv5s is proposed. Firstly, to solve the problems arising from the anchor frame generated by the network through K-means clustering, the dynamic anchor frame structure DAFS is introduced in the input stage. Secondly, the SPP-D (Spatial Pyramid Pooling-Defect) improved fr… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
3
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
4
1
1

Relationship

1
5

Authors

Journals

citations
Cited by 6 publications
(3 citation statements)
references
References 23 publications
0
3
0
Order By: Relevance
“…Some researchers have proposed new structures to enhance the feature extraction capabilities of the network by reducing the loss of detailed information. For example, He G et al [ 29 ] introduced the SPP-D module, which reduces the feature information loss caused by max-pooling in the SPP module by enhancing the reuse of feature information, thereby improving small object detection capabilities. Sunkara R et al [ 30 ] utilized the SPD-Conv structure to replace the original downsampling structure, successfully reducing information loss and improving the algorithm’s capability to detect small objects.…”
Section: Related Workmentioning
confidence: 99%
See 1 more Smart Citation
“…Some researchers have proposed new structures to enhance the feature extraction capabilities of the network by reducing the loss of detailed information. For example, He G et al [ 29 ] introduced the SPP-D module, which reduces the feature information loss caused by max-pooling in the SPP module by enhancing the reuse of feature information, thereby improving small object detection capabilities. Sunkara R et al [ 30 ] utilized the SPD-Conv structure to replace the original downsampling structure, successfully reducing information loss and improving the algorithm’s capability to detect small objects.…”
Section: Related Workmentioning
confidence: 99%
“…Many researchers have proposed their own improvement methods for this issue. For example, He G et al [ 29 ] used DIoU-NMS to replace the original NMS, successfully improving the detection accuracy of small objects in complex backgrounds. Zhang H et al [ 36 ] combined Cluster-NMS [ 37 ] with DIoU-NMS to propose the Cluster-DIoU-NMS method, which solves both over-suppression and overlapping issues.…”
Section: Related Workmentioning
confidence: 99%
“…Moreover, the surface defects in camera modules are small objects, and the complex industrial contexts such as camera module fabrication demand high accuracy and real‐time target recognition. [ 13 ] Therefore, a target recognition algorithm that can detect small targets quickly and accurately is needed for camera module surface defect detection. However, the YOLO series target detection algorithm, despite its high detection efficiency, has difficulties in detecting small targets.…”
Section: Introductionmentioning
confidence: 99%