ABSTRACT
In response to the challenges posed by low detection accuracy resulting from a wide range of surface defects, intricate textures, and minute defect targets in steel surfaces, this paper introduces an innovative defect detection model called DCS-YOLOv8, which builds upon the foundation of YOLOv8. Firstly, Real-ESRGAN (Real-Enhanced Super-Resolution GAN) is used to enhance image resolution, effectively addressing the challenge of identifying minuscule defects within the dataset. Furthermore, DCN (Deformable Convolutions) are seamlessly integrated into the backbone network to amplify the network's capability for multi-scale feature extraction, which empowers the network to adeptly navigate intricate background information and concentrate on pinpointing target objects. Lastly, to tackle the issues of elevated false negative rates and diminished detection precision, this paper designs a module based on the CBAM (Concentration-Based Attention Module) and SCSE (Concurrent Spatial and Channel Squeeze and Excitation) attention modules. It facilitates comprehensive information acquisition, enriches the fusion of channel and spatial features, and elevates feature map expression. Regarding experimental outcomes, the enhanced YOLOv8 algorithm shows outstanding detection performance, achieving 78.6% mAP on the NEU-DET dataset, which marks a 4.4% enhancement over the original YOLOv8 network. Notably, the algorithm attains a detection speed of 143 FPS. When juxtaposed with other prominent object detection algorithms, it unequivocally affirms the efficacy and supremacy of this approach, underscoring its potential significance in industrial applications.
- Xiong Z, Li Q, Mao Q, A 3D laser profiling system for rail surface defect detection. Sensors, 2017, 17(8): 1791.Google ScholarCross Ref
- Liang Y, Xu K, Zhou P. Mask gradient response-based threshold segmentation for surface defect detection of milled aluminum ingot. Sensors, 2020, 20(16): 4519.Google ScholarCross Ref
- Yefan Zhou, "Research on Image-Based Automatic Wafer Surface Defect Detection Algorithm," Journal of Image and Graphics, Vol. 7, No. 1, pp. 26-31, March 2019. doi: 10.18178/joig.7.1.26-31Google ScholarCross Ref
- Teow Wee Teo and Mohd Zaid Abdullah, "Solar Cell Micro-Crack Detection Using Localised Texture Analysis," Journal of Image and Graphics, Vol. 6, No. 1, pp. 54-58, June 2018. doi: 10.18178/joig.6.1.54-58Google ScholarCross Ref
- Girshick R, Donahue J, Darrell T, Rich feature hierarchies for accurate object detection and semantic segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition. 2014: 580-587.Google Scholar
- Ren S, He K, Girshick R, Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in neural information processing systems, 2015,28: 1440-1448.Google Scholar
- He K, Gkioxari G, Dollár P, Mask R-CNN. Proceedings of the IEEE international conference on computer vision. 2017: 2961-2969.Google Scholar
- Zhao W, Chen F, Huang H, A new steel defect detection algorithm based on deep learning. Computational Intelligence and Neuroscience, 2021, 2021: 1-13.Google ScholarDigital Library
- Liu W, Anguelov D, Erhan D, Ssd: Single shot multibox detector. Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14. Springer International Publishing, 2016: 21-37.Google Scholar
- Redmon J, Farhadi A. Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767, 2018.Google Scholar
- Redmon J, Divvala S, Girshick R, You only look once: Unified, real-time object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 779-788.Google Scholar
- Redmon J, Farhadi A. YOLO9000: better, faster, stronger. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 7263-7271.Google Scholar
- Bochkovskiy A, Wang C Y, Liao H Y M. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934, 2020.Google Scholar
- Li C, Li L, Jiang H, YOLOv6: A single-stage object detection framework for industrial applications. arXiv preprint arXiv:2209.02976, 2022.Google Scholar
- Wang C Y, Bochkovskiy A, Liao H Y M. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023: 7464-7475.Google Scholar
- Junjie X, Minping J, Feiyun X, A method for workpiece surface small-defect detection based on CutMix and YOLOv3. Journal of Southeast University (English Edition), 2021, 37(2): 128-136.Google Scholar
- Zhao, H., Wan, F., Lei, G., Xiong, Y., Xu, L., Xu, C., & Zhou, W. LSD-YOLOv5: A Steel Strip Surface Defect Detection Algorithm Based on Lightweight Network and Enhanced Feature Fusion Mode. Sensors, 2023, 23(14), 6558.Google Scholar
- Zhu X, Hu H, Lin S, Deformable convnets v2: More deformable, better results. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019: 9308-9316.Google Scholar
- Bolya D, Zhou C, Xiao F, Yolact: Real-time instance segmentation. Proceedings of the IEEE/CVF international conference on computer vision. 2019: 9157-9166.Google Scholar
- Wang C Y, Liao H Y M, Wu Y H, CSPNet: A new backbone that can enhance learning capability of CNN. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops. 2020: 390-391.Google Scholar
- Lin T Y, Dollár P, Girshick R, Feature pyramid networks for object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2117-2125.Google Scholar
- Liu S, Qi L, Qin H, Path aggregation network for instance segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 8759-8768.Google Scholar
- Ge Z, Liu S, Wang F, Yolox: Exceeding yolo series in 2021. arXiv preprint arXiv:2107.08430, 2021.Google Scholar
- Li X, Wang W, Wu L, Generalized focal loss: Learning qualified and distributed bounding boxes for dense object detection. Advances in Neural Information Processing Systems, 2020, 33: 21002-21012.Google Scholar
- Cubuk E D, Zoph B, Mane D, Autoaugment: Learning augmentation strategies from data. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019: 113-123.Google Scholar
- Wang X, Xie L, Dong C, Real-ESRGAN: Training real-world blind super-resolution with pure synthetic data. Proceedings of the IEEE/CVF international conference on computer vision. 2021: 1905-1914.Google Scholar
- Lin T Y, Maire M, Belongie S, Microsoft coco: Common objects in context. Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13. Springer International Publishing, 2014: 740-755.Google Scholar
- Woo S, Park J, Lee J Y, Cbam: Convolutional block attention module. Proceedings of the European conference on computer vision (ECCV). 2018: 3-19.Google Scholar
- Roy A G, Navab N, Wachinger C. Concurrent spatial and channel ‘squeeze & excitation’ in fully convolutional networks. Medical Image Computing and Computer Assisted Intervention–MICCAI 2018: 21st International Conference, Granada, Spain, September 16-20, 2018, Proceedings, Part I. Springer International Publishing, 2018: 421-429.Google Scholar
- Song K, Yan Y. A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects. Applied Surface Science, 2013, 285: 858-864.Google ScholarCross Ref
Index Terms
- DCS-YOLOv8: An Improved Steel Surface Defect Detection Algorithm Based on YOLOv8
Recommendations
An Improved YOLOv5 with Structural Reparameterization for Surface Defect Detection
Artificial Neural Networks and Machine Learning – ICANN 2023AbstractSurface defects produced by the manufacturing process directly degrades the quality of industrial materials such as hot-rolled steel. However, existing methods for detecting surface defects cannot meet the requirements in terms of speed and ...
YOLO-IMF: An Improved YOLOv8 Algorithm for Surface Defect Detection in Industrial Manufacturing Field
Metaverse – METAVERSE 2023AbstractSurface quality is an effective metric to evaluate the quality of industrial products. The traditional automatic detection methods such as eddy current detection method, infrared detection method and magnetic flux leakage method are limited to ...
Defect Detection of Production Surface Based on CNN
Green, Pervasive, and Cloud ComputingAbstractWith the continuous development of artificial intelligence, great progress has been made in the field of object detection. Defect detection is a branch of the field of object detection, as long as the purpose is to locate and classify defects on ...
Comments