计算机科学 ›› 2023, Vol. 50 ›› Issue (11A): 230200150-6.doi: 10.11896/jsjkx.230200150
邹宸玮, 么娆
ZOU Chenwei, YAO Rao
摘要： 超声相控阵检测是一种常用的无损检测(NDT)技术,用于工件缺陷检测与评估。为实现现代工业大数据化和自动化,解决缺陷检测过程中产生的超声相控阵图像信息缺失和散斑噪声问题,并实现对各类型缺陷的准确识别,提出了一种基于机器视觉的缺陷识别方法,该方法在利用改进PM微分方程对图像进行去噪处理后,提取图像特征作为粒子群优化的BP神经网络的实验数据。实验结果表明,所提方法的训练集精准率为99.543%,相比传统的BP网络模型提高了1.833%,能够在准确实现缺陷识别的同时,保持模型良好的性能。
中图分类号:
| [1]LIU H,CHEN Z L,ZHANG L.A machine learning-basedmethod for detecting and classifying tool surface defects[J].Computer Measurement and Control,2021,29(6):64-68. [2]WANG S.Solder defect detection method based on machine vision and deep learning [D].Guangdong:Guangdong University of Technology,2022. [3]ZHANG M.Research on LED surface defect detection technology based on machine learning [D].Hefei:Hefei University of Technology,2021. [4]ZHANG X L,LI J H,LI J X,et al.Hybrid optimized RBF-BP network for plate shape defect identification[J].Fuzzy Systems and Mathematics,2020,34(1):149-156. [5]LIU E D.Pipeline defect identification based on BP neural network [J].Electronic Technology and Software Engineering,2017(2):11-12. [6]ZHU L Z,CHENM Y.A research on a machine learning based automotive glue coating defect detection[J].Journal of Shenyang University of Technology,2018,37(4):18-22. [7]MCNAB A,CAMPBELL M J.Ultrasonic phased arrays for nondestructive testing[J].NDT International,1987,20(6):333-337. [8]THRANE L,YURA H T,ANDERSEN P E.Analysis of optical coherence tomography systems based on the extended Huygens－－Fresnel principle[J].JOSA A,2000,17(3):484-490. [9]LIU Z Y.Application of partial differential equations in image denoising [D].Changchun:Jilin University,2018. [10]FANG Z,HU X H,CHEN Y.Anisotropic diffusion filtering al- gorithm based on multi-directional median filtering[J].Compu-ter Engineering and Applications,2017,53(4):195-199. [11]FU L J,YAO Y,FU Z L.A medical image filtering method combining median filtering and anisotropic diffusion[J].Computer Applications,2014,34(1):145-148. [12]WAN S,LI L M,HUANG Y Q.An image denoising model incorporating partial differential equations and median filtering[J].Computer Applications,2011,31(9):2512-2514. [13]CHEN T,ZHANG Q G,LIU A.Implementation of SVM classification for graphical texture detection and welding defects based on grayscale co-occurrence matrix[J].Journal of Luoyang Institute of Technology (Natural Science Edition),2022,32(1):53-61,67. [14]LIU G Y,HUANG Y,CAO Y,et al.Research on image texture feature extraction based on grayscale co-occurrence matrix[J].Science and Technology Wind,2021(12):61-64. [15]MOU H B,WANG S W,QI D W,et al.Wood defect recognition based on grayscale covariance matrix and fuzzy BP neural network[J].Forest Engineering,2017,33(4):40-43,54. [16]YANG C X,ZHANG J,TAO Z Y,et al.Metal surface defects classification based on PSO improved BP neural network algorithm[J].Manufacturing Automation,2018,40(7):5-7. [17]LIY X.Research on potato external defects detection method based on PSO-BP neural network [D].Jinan:Shandong University of Technology,2021. [18]DAS H,JENA A K,NAYAK J,et al.A novel PSO based back propagation learning-MLP (PSO-BP-MLP) for classification[M]//Computational intelligence in data mining-volume 2.Springer,New Delhi,2015:461-471. [19]SANG J G.Research on pump fault diagnosis based on pso-bp neural network algorithm[C]//2019 IEEE 8th Joint International Information Technology and Artificial Intelligence Conference(ITAIC).IEEE,2019. |
|
||
首页