基于超分辨率改进Faster R-CNN的点阵结构内部缺陷判识方法

doi:10.3901/JME.2022.21.266

机械工程学报 ›› 2022, Vol. 58 ›› Issue (21): 266-273.doi: 10.3901/JME.2022.21.266

扫码分享

基于超分辨率改进Faster R-CNN的点阵结构内部缺陷判识方法

温银堂^1,2, 付凯^1,2, 张玉燕^1,2, 张芝威^1,2

1. 燕山大学电气工程学院秦皇岛 066004;
2. 燕山大学测试计量技术及仪器河北省重点实验室秦皇岛 066004

收稿日期:2021-11-18 修回日期:2022-04-08 出版日期:2022-11-05 发布日期:2022-12-23
通讯作者: 张玉燕(通信作者),女,1976年出生,博士,教授,博士研究生导师。主要研究方向为动态测量与分析、结构健康监测、电容层析成像。E-mail:yyzhang@ysu.edu.cn
作者简介:温银堂,男,1978年出生,博士,研究员,博士研究生导师。主要研究方向为智能检测与评估技术。E-mail:ytwen@ysu.edu.cn
基金资助:
河北省自然科学基金（E2017203240）和河北省科技计划（20310401D）资助项目。

A Method of Improved Faster R-CNN for Detecting Internal Defect of Metal Lattice Structure Based on Super-resolution Reconstruction

WEN Yintang^1,2, FU Kai^1,2, ZHANG Yuyan^1,2, ZHANG Zhiwei^1,2

1. School of Electrical Engineering, Yanshan University, Qinhuangdao 066004;
2. Key Lab of Hebei Measurement Technology and Instrumentation of Hebei Province, Yanshan University, Qinhuangdao 066004

Received:2021-11-18 Revised:2022-04-08 Online:2022-11-05 Published:2022-12-23

摘要/Abstract

摘要： 利用增材制造技术制备金属三维点阵结构件过程中，结构内部经常会出黏连、断裂等多种细小缺陷，导致样件结构功能下降。根据缺陷结构与正常结构之间的特征区别，提出了一种针对金属点阵结构内部出现的细小缺陷自动判识的方法。利用X-射线微聚焦CT扫描金属点阵结构获得原始输入图片，在Faster R-CNN （Faster region-based convolutional neural networks）框架的基础上，改进原有特征提取网络，开发图像超分辨率重建模块。通过对工业CT图片的局部细节特征增强，实现了快速有效地识别细小缺陷的类型，以及缺陷位置信息的标注。试验证明，改进Faster R-CNN模型对金属点阵结构样件内部的两种典型细小缺陷识别的平均正确率高达93.5%。研究结果表明，通过超分辨率网络对图像进行放大，可以提高细小缺陷的特征提取，通过加深网络加强特征学习，从而实现了点阵结构内部细小缺陷的自动判识。

关键词: 金属点阵结构, 缺陷识别, CT切片图像, 改进Faster R-CNN, 超分辨率重建

Abstract: In the process of fabricating metal 3D lattice structure with additive manufacturing technology, many small defects such as adhesion and fracture often occur in the structure, which leads to the structural function decline of the sample. According to the characteristic difference between defective structure and normal structure, an automatic identification method for small defects in metal lattice structure is proposed. The original input images are obtained by scanning metal lattice structures with X-ray microfocusing CT, and the original feature extraction network is improved based on the Faster R-CNN(Faster region-based convolutional neural networks) framework. The image super-resolution reconstruction module is developed. By enhancing the local detail features of industrial CT images, it can quickly and effectively identify the types of small defects and mark the location information of defects. Experimental results show that the improved Faster R-CNN model has an average correct rate of 93.5% for identifying two typical small defects in metal lattice structure samples. The results show that the feature extraction of small defects can be improved by using super-resolution network to enlarge the image, and the feature learning can be enhanced by deepening the network, thus realizing the automatic identification of small defects in the lattice structure.

Key words: metal lattice structure, defect identification, CT slices, improved Faster R-CNN, super-resolution reconstruction

中图分类号:

TG115

温银堂, 付凯, 张玉燕, 张芝威. 基于超分辨率改进Faster R-CNN的点阵结构内部缺陷判识方法[J]. 机械工程学报, 2022, 58(21): 266-273.

WEN Yintang, FU Kai, ZHANG Yuyan, ZHANG Zhiwei. A Method of Improved Faster R-CNN for Detecting Internal Defect of Metal Lattice Structure Based on Super-resolution Reconstruction[J]. Journal of Mechanical Engineering, 2022, 58(21): 266-273.

参考文献

[1] MOON S K, TAN Y E, JIHONG H, et al. Application of 3D printing technology for designing light-weight unmanned aerial vehicle wing structures[J]. International Journal of Precision Engineering and Manufacturing -Green Technology, 2014, 1(3):223-228.
[2] XIONG Jian, MA Li, WU Linzhi, et al. Fabrication and crushing behavior of low density carbon fiber composite pyramidal truss structures[J]. Composite Structures, 2010, 92(11):2695-2702.
[3] SHIDID D, LEARY M, CHOOG P, et al. Just-in-time design and additive manufacture of patient-specific medical implants[J]. Physics Procedia, 2016, 83:4-14.
[4] ZHOU Hao, ZHANG Xiaoyu, ZENG Huizhong, et al. Lightweight structure of a phase-change thermal controller based on lattice cells manufactured by SLM[J]. Chinese Journal of Aeronautics, 2019, 32(7):1727-1732.
[5] LIANG Hao, RAYMONT David, YAN Chunze, et al. Design and additive manufacturing of cellular lattice structures[C]//The International Conference on Advanced Research in Virtual and Rapid Prototyping (VRAP). Taylor & Francis Group, Leiria. 2011:249-254.
[6] LIU Fei, ZHANG David, ZHANG Peng, et al. Mechanical properties of optimized diamond lattice structure for bone scaffolds fabricated via selective laser melting[J]. Materials, 2018, 11(3):374.
[7] 郑聃, 李瑞迪, 宋波, 等. NiTi气雾化制粉工艺对选区激光熔化成型性、制件超弹性的影响[J]. 机械工程学报, 2020, 56(15):104-109. ZHENG Dan, LI Ruidi, SONG Bo, et al. Effect of NiTi powder gas atomization process on the selective laser melting moldability and alloys' superelastic[J]. Journal of Mechanical Engineering, 2020, 56(15):104-109.
[8] SING S L, WIRIA F E, YEONG W Y. Selective laser melting of lattice structures:A statistical approach to manu-facturability and mechanical behavior[J]. Robotics and Computer-Integrated Manufacturing, 2018, 49:170-180.
[9] MAZUR M, LEARY M, SUN Shoujin, et al. Deformation and failure behaviour of Ti-6Al-4V lattice structures manufactured by selective laser melting (SLM)[J]. The International Journal of Advanced Manufacturing Technology, 2016, 84(5-8):1391-1411.
[10] 邬冠华, 熊鸿建. 中国射线检测技术现状及研究进展[J]. 仪器仪表学报, 2016, 37(8):1683-1695. WU Guanhua, XIONG Hongjian. Radiography testing in China[J]. Chinese Journal of Scientific Instrument, 2016, 37(8):1683-1695.
[11] 姬文苏, 丁玉奎. 火箭发动机多层粘接结构超声检测研究[J]. 兵工学报, 2015, 36(S1):364-368. JI Wensu, DING Yukui. Study of ultrasonic detection for multi-layer adhesive joints of solid propellant rocket engines[J]. Acta Armamentarii, 2015, 36(S1):364-368.
[12] 郭健, 张丹, 马国义, 等. 无损检测(NDT)——磁粉检测(MT)技术[J]. 工程与试验, 2011, 51(3):55-58. GUO Jian, ZHANG Dan, MA Guoyi, et al. Nondestructive testing-magnetic particle testing technology[J]. Engineering & Test, 2011, 51(3):55-58.
[13] 杨理践, 耿浩, 高松巍. 长输油气管道漏磁内检测技术[J]. 仪器仪表学报, 2016, 37(8):1736-1746. YANG Lijian, GENG Hao, GAO Songwei. Magnetic flux leakage internal detection technology of the long distance oil pipeline[J]. Chinese Journal of Scientific Instrument, 2016, 37(8):1736-1746.
[14] GARCIA-MARTIN J, GOMEZ-GIL J, VAZQEZ-SANCHEZ E. Non-destructive techniques based on eddy current testing[J]. Sensors, 2011, 11(3):2525.
[15] KRAUSS H, ZEUGNER T, ZAEH M F. Layerwise monitoring of the selective laser melting process by thermography[J]. Physics Procedia, 2014, 56:64-71.
[16] NOURI H, GUEDDASMA S, BELHABIB S. Structural im-perfections in additive manufacturing perceived from the X-ray micro-tomography perspective[J]. Journal of Materials Processing Technology, 2016, 234:113-124.
[17] SANDE K, UIJLINGS J, GEVERS T, et al. Segmentation as selective search for object recognition[C]//International Conference on Computer Vision. Barcelona, Spain:IEEE Computer Society, 2011:1879-1886.
[18] SERMANET P, EIGEN D, ZHANG Xiang, et al. Overfeat:Integrated recognition, localization and detection using convolutional networks[C]//2nd International Conference on Learning Representations, ICLR 2014. Banff, Canada, Université de Montreal, 2014.
[19] REN Shaoqing, HE Kaiming, GIRSHICK Ross, et al. Faster R-CNN:Towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 39(6):1137-1149.
[20] CAI Zhaowei, FAN Quanfu, FERIS R, et al. A unified multi-scale deep convolutional neural network for fast object detection[J]. Lecture Notes in Computer Science, 2016, 354-370.
[21] JOSEPH R, ALI F. Yolov3:An incremental improvement[EB/OL]. arXiv preprint[2020-2-25]. https://doi.org/10.48550/arXiv.1804.02767.
[22] 田珠, 桂志国, 张鹏程, 等. Faster_RCNN用于工业火花塞图像焊缝缺陷检测[J]. 测试技术学报, 2020, 34(1):34-40. TIAN Zhu, GUI Zhiguo, ZHANG Pengcheng, et al. Faster_RCNN for industrial spark plug image weld defect inspection[J]. Journal of Test and Measurement Technology, 2020, 34(1):34-40.
[23] 梁杰, 李磊, 任君, 等. 基于深度学习的红外图像遮挡干扰检测方法[J]. 兵工学报, 2019, 40(7):1401-1410. LIANG Jie, LI Lei, REN Jun, et al. Infrared image occlusion interference detection method based on deep learning[J]. Acta Armamentarii, 2019, 40(7):1401-1410.
[24] XIE Lele, AHMAD T, JIN Lianwen, et al. A new CNN-based method for multi-directional car license plate detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(2):507-517.
[25] 张玉燕, 李永保, 温银堂, 等. 基于Faster R-卷积神经网络的金属点阵结构缺陷识别方法[J]. 兵工学报, 2019, 40(11):2329-2335. ZHANG Yuyan, LI Yongbao, WEN Yintang, et al. Internal defect detection of metal three-dimensional multi-layer lattice structure based on Faster R-CNN[J]. Acta Armamentarii, 2019, 40(11):2329-2335.
[26] ZOU Zhengxia, SHI Zhenwei, GUO Yuhong, et al. Object detection in 20 years:A survey[EB/OL]. arX-iv preprint[2019-05-16] https://doi.org/10.48550/arXiv.1905.05055.
[27] DONG Chao, LOY Chenchang, HE Kaiming, et al. Learning a deep convolutional network for image super-resolution[C]//European conference on computer vision. Springer, Cham, 2014:184-199.

基于超分辨率改进Faster R-CNN的点阵结构内部缺陷判识方法

A Method of Improved Faster R-CNN for Detecting Internal Defect of Metal Lattice Structure Based on Super-resolution Reconstruction

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 12

编辑推荐

Metrics

本文评价

[1]	闫守鑫, 王伟, 苏鹏飞, 贠超. 大型复杂锻件打磨路径的智能化规划方法[J]. 机械工程学报, 2024, 60(11): 216-225.
[2]	朵文博, 李宏伟, 李帅兵, 杨栋, 卢保朋, 康永强, 曹炳磊. 基于太赫兹成像的层状复合绝缘结构内部分层缺陷SSA-CNN定量表征^*[J]. 电气工程学报, 2023, 18(3): 63-72.
[3]	姜洪权, 贺帅, 高建民, 王荣喜, 高智勇, 王晓桥, 夏锋社, 程雷. 一种改进卷积神经网络模型的焊缝缺陷识别方法[J]. 机械工程学报, 2020, 56(8): 235-242.
[4]	刘菲菲, 刘松平, 傅天航, 李乐刚, 白金鹏, 史俊伟. 复合材料结构中R区的超声波反射行为及其检测应用[J]. 机械工程学报, 2017, 53(12): 35-43.
[5]	盛朝阳;刚铁;迟大钊. 基于分水岭方法的超声TOFD检测图像分割[J]. , 2011, 47(8): 35-40.
[6]	胡祥超;罗飞路;何赟泽;唐莺. 脉冲交变磁场测量技术缺陷识别与定量评估[J]. , 2011, 47(4): 17-22.
[7]	迟大钊;刚铁;姚英学. 基于非线性小波收缩的超声信号缺陷识别方法[J]. , 2011, 47(18): 1-6.
[8]	车红昆;吕福在;项占琴. 多特征SVM-DS融合决策的缺陷识别[J]. , 2010, 46(16): 101-105.
[9]	李大勇;高桂丽;董静薇. 非线性声学和时间反转声学在材料缺陷识别中的应用现状评述[J]. , 2009, 45(1): 1-8.
[10]	宗培;曹雷;邵国良;彭飞. 焊接结构质量的主成分分析[J]. , 2005, 41(5): 65-68.
[11]	刘旭;夏金东;弓乐;吴淼;孙智. 射频信号在超声检测缺陷识别中的应用研究[J]. , 2002, 38(4): 84-87.
[12]	刚铁;吴林. 超声检测中的多源信息融合技术与缺陷识别[J]. , 1999, 35(1): 11-14.