A Method for Detecting Internal Defects of Metal Lattice Structure Based on YOLOv3 Algorithm
ZHANG Yu-yan1,2, REN Teng-fei1,2, WEN Yin-tang1,2
1. School of Electrical Engineering, Yanshan University, Qinhuangdao, Hebei 066004, China
2. Hebei Province Key Laboratory of Measuring and Testing Technologies and Instruments,Yanshan University, Qinhuangdao, Hebei 066004, China
Abstract To solve the problem that the defect in the lattice structure is difficult to accurately identify due to the small size and weak feature,an intelligent defect recognition method based on YOLOv3 algorithm is proposed.This method takes advantage of the deep learning network model in feature extraction,uses a multi-scale network to predict and treats the classification and location of defects as regression problems.The proposed algorithm realizes the identification of internal defects in a 3D printed lattice structure.And the detect recall is 96.6%,the accuracy is 93.2%,and the mean average precision value of model is 0.957.It provides a basis for further accurate characterization of defects and analysis of the effects of defects on the performance of lattice structures.
ZHANG Yu-yan,REN Teng-fei,WEN Yin-tang. A Method for Detecting Internal Defects of Metal Lattice Structure Based on YOLOv3 Algorithm[J]. Acta Metrologica Sinica, 2022, 43(1): 7-13.
［1］ Thompson A, Maskery I, Leach R K. X-ray computed tomography for additive manufacturing: a review ［J］. Measurement Science & Technology, 2016, 27 (7): 072001.
［2］ Kang D, Park S, Son Y, et al. Multi-lattice inner structures for high-strength and lightweight in metal selec-tive laser melting process ［J］. Materials and Design, 2019, 175 (10): 77-86.
［3］ Valavanis I, Kosmopoulos D. Multiclass defect detection and classification in weld radiographic images using geo-metric and texture features ［J］. Expert Systems With Applications, 2010, 37 (12): 7606-7614.
［4］ 胡巍, 刘志, 孙广玲, 等. 基于频域特征的焊缝X射线图像缺陷检测 ［J］. 工业控制计算机, 2017, 30 (10): 78-79+81.
Hu W, Liu Z, Sun G L, et al. Defect Detection in X-ray Image of Weld Seam Based on Frequency Feature ［J］. Industrial Control Computer,2017,30 (10):78-79+81.
［5］ 程淑红, 马继勇, 张仕军, 等. 改进的混合高斯与YOLOv2融合烟雾检测算法 ［J］. 计量学报, 2019, 40 (5): 798-803.
Cheng S H, Ma J Y, Zhang S J, et al. Smoke Detection Algorithm Combined with Improved Gaussian Mixture and YOLOv2 ［J］. Acta Metrologica Sinic, 2019, 40 (5): 798-803.
［6］ 张世辉,耿勇,张笑维. 基于深度图像利用BP网络实现遮挡边界检测［J］. 计量学报, 2020, 41 (10): 1205-1211.
Zhang S H, Geng Y, Zhang X W. Using BP Network for Occlusion Boundary Detection Based on Depth Imageg［J］.Acta Metrologica Sinic,2020,41(10):1205-1211.
［7］ Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation［C］//Proceedings of the IEEE conference on computer vision and pattern recognition. 2014: 580-587.
［8］ Girshick R. Fast R-CNN［C］//Proceedings of the IEEE international conference on computer vision, 2015: 1440-1448.
［9］ Ren S Q, He K M, Girshick R, et al. Faster R-CNN: Towards realtime object detection with region proposal networks［C］//Advances in neural information processing systems. 2015: 91-99.
［10］ Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection［C］//Proce-edings of the IEEE conference on computer vision and pattern recognition. 2016: 779-788.
［11］ Redmon J, Farhadi A. YOLO9000: better, faster, stronger［C］//Proceeding of the IEEE conference on computer vision and pattern recognition. 2017: 7263-7271.
［12］ Redmon J, Farhadi A. YOLOv3: An incremental imp-rovement［C］//Conference on Computer Vision and Pattern. Salt Lake City: IEEE, 2018: 1-4.
［13］ He K, Zhang X, Ren S, et al. Deep Residual Learning for Image Recognition［C］//IEEE Conference on Com-puter Vision & Pattern Recognition. IEEE Computer Society, 2016.
［14］ Du Y X, Gu D D, Xi L X, et al. Laser additive man-ufacturing of bio-inspired lattice structure: Forming quality, microstructure and energy absorption behavior ［J］. Materials Science and Engineering A, 2020, 773: 138857.
［15］ 温银堂,高亭亭,张玉燕. 3D打印复杂点阵结构的缺陷三维可视化方法［J］. 计量学报, 2020, 41 (9): 1077-1081.
Wen Y T, Gao T T, Zhang Y Y. 3D Visualization Method for Complex Lattice Structure Defects in 3D Printing［J］. Acta Metrologica Sinic, 2020, 41 (9): 1077-1081.
［16］ 顾冬冬, 张红梅, 陈洪宇等. 航空航天高性能金属材料构件激光增材制造 ［J］. 中国激光, 2020, 47 (5): 32-55.
Gu D D, Zhang H M, Zhang H Y, et al. Laser Additive Manufacturing of High-Performance Metallic Aerospace Components ［J］. Chinese Journal of Lasers, 2020, 47 (5): 32-55.
［17］ 谭芳,穆平安,马忠雪. 基于YOLOv3检测和特征点匹配的多目标跟踪算法［J］. 计量学报, 2021, 42 (2): 157-162.
Tan F, Mu P A, Ma Z X. Multi-target Tracking Algorithm Based on YOLOv3 Detection and Feature Point Matching［J］. Acta Metrologica Sinic, 2021, 42 (2): 157-162.
2022, Vol. 43 
