Scratch defect detection model on wooden board surface with complex texture

被引：0
|
作者：
Hu Q. [1 ]
Qin W. [1 ]
Liu C. [1 ]
Shi W. [2 ]
机构：
[1] School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai
[2] China Mobile (Shanghai) ICT Co., Ltd., Shanghai
来源：
Jisuanji Jicheng Zhizao Xitong/Computer Integrated Manufacturing Systems, CIMS | 2024年 / 30卷 / 01期
关键词：
deformable convolution; Faster RCNN; regression loss; rotating bounding box; scratch detection;
D O I：
10.13196/j.cims.2021.0513
中图分类号：
学科分类号：
摘要：
To improve the automation level of the wood processing production line, a scratch defect detection model on the wood surface based on Faster RCNN was proposed to identify and locate scratch defects under different texture backgrounds. In the image preprocessing stage, an improved bilateral filtering algorithm was proposed to smooth the texture background while maintaining the details of the scratches. A gray-scale adaptive scratch generation method was proposed for data enhancement. The deformable convolution was introduced to enhance the feature extraction ability of the model, and the rotating bounding box was used and a new bounding box regression loss function was proposed to solve the problem that the proportion of scratch defects in the horizontal bounding box was much smaller than the texture background. The images collected by the actual wood board processing production line verified the effectiveness of the proposed model. The proposed model was compared with other defect detection methods, and the results proved the superiority of the proposed model. © 2024 CIMS. All rights reserved.
引用
收藏
页码：78 / 89
页数：11
相关论文
共 39 条
  • [1] LI Yiting, XIE Qingsheng, HUANG Haisong, Surface defect detection based on fast regions wth convoiutionai neutral net work, Computer Integrated Manufacturing Systems, 25, 8, pp. 1897-1907, (2019)
  • [2] KRIZHEVSKY A, SUTSKEVER I, HINTON G E., ImageNet 11X1110" wHi deep convoiutionai neural networks, Advances in Neural Information Processing Systems, 25, pp. 1097-1105, (2012)
  • [3] GIRSHICK R, DONAHUE J, DARRELL T, Et al., Rch feature hierarchies for accurate object detection and semantic segmentatOin, Proceeding"s of the IEEE Conference inn Gmi-puter Vssiinn and Pattern Recognition, pp. 580-587, (2014)
  • [4] GIRSHICK R., Fast R-CNN, Proceedings of the IEEE In-ternationai Conference inn Computer Vssiinn, pp. 1440-1448, (2015)
  • [5] REN S Q, HE K M, GIRSHICK R, Et al., Faster R-CNN: Towards rea-time object detection wHi region proposai net works
  • [6] CHA Y J, CHOI W, SUH G, Et al., Autonomous structurai vtsuai inspection using" region-based deep learning" for detecting mukiple damage types, Computer Aided Civii and Infrastructure Eng"ineering, 33, 9, pp. 731-747, (2018)
  • [7] ZHONG J, LIU Z, HAN Z, Et al., A CNN-based defect inspection method for catenary spit pins in highspeed raiiway, IEEE Transactions on Instrumentation and Measurement, 68, 8, pp. 2849-2860, (2018)
  • [8] TAO X, ZHANG D, WANG Z, Et al., Detection of power line insuLatirr defects using aeriai images anaiyzed wtth convoiution ai neurai networks, IEEE Transactions on Systems, Man, and Cybernetics: Systems, 50, 4, pp. 1486-1498, (2018)
  • [9] XUE Y, LI Y., A fast detection method via region-based fuiiy convoiutionai neurai networks for slieid tunnei iinmg" defects, Computer-Added Civii and Infrastructure Engineeting, 33, 8, pp. 638-654, (2018)
  • [10] DING R, DAI L, LI G, Et al., TDD-net: A tiny defect detection network for printed circmt boards, CAAI Transac tions on Inteiiigence Technoiogy, 4, 2, pp. 110-116, (2019)
1234