轨道缺陷图像检测算法的优化  

Optimization of image detection algorithm for track defect

作　　者：闵锐[1] 黄楚云[1] 张晓峻 孙晶华[2] MIN Rui;HUANG Chu-yun;ZHANG Xiao-jun;SUN Jing-hua(School of Science»Hubei University of Technology,Wuhan 430068,China;College of Physics and Optoelectronic Engineering,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]湖北工业大学理学院,湖北武汉430068 [2]哈尔滨工程大学物理与光电工程学院,黑龙江哈尔滨150001

出　　处：《物理实验》2022年第2期9-15,共7页Physics Experimentation

基　　金：国家自然科学基金面上项目(No.61377084)。

摘　　要：根据轨道健康状况无损检测的需求,基于OpenCV软件对轨道缺陷图像检测BM3D算法进行了优化.利用小波变换优化BM3D算法,提出BM2D+1D的算法模式,改进后算法的运算效率提高了39.6%,图像峰值信噪比提高了0.388dB.利用图像增强LOG算子和Gamma算子对Canny算法进行改进,不仅优化了算法运算复杂度,同时滤除了大部分虚假边缘信息和无意义像素信息,运算效率提高了81.4%,使Canny算法的阈值可大范围调节并确定最优解,提高了算法的灵活度.According to the requirements of track health non-destructive testing,the track defect detection BM3D algorithm was optimized based on OpenCV software.Wavelet transform was used to optimize BM3D algorithm,and the algorithm mode of BM2D+1D was proposed.The operation efficiency was improved by 39.6%9 and the image peak signal-to-noise ratio was improved by 0.388 dB.Canny algorithm was improved by using image enhancement operators,LOG operator and Gamma operator,which not only optimized the time complexity of the algorithm,but also filtered out most of the false edge information and meaningless pixel information.The operation efficiency was improved by 81.4%.The threshold of Canny algorithm could be adjusted in a large range to determine the optimal solution,which improved the flexibility of the algorithm.

关 键 词：轨道缺陷检测 BM3D算法 LOG算子 GAMMA算子 CANNY算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] U216.3[自动化与计算机技术—计算机科学与技术]