基于改进LOG算子与Otsu算法结合的物体表面图像残损检测方法  

Image Damage Detection Method of Object Surface Based on Improved LOG Operator and Otsu Algorithm

作　　者：李相格[1] LI Xiangge(China Railway First Survey and Design Institute Group Co.,Ltd.,Xi'an 710043,China)

机构地区：[1]中铁第一勘察设计院集团有限公司,西安710043

出　　处：《兰州交通大学学报》2024年第1期59-63,共5页Journal of Lanzhou Jiaotong University

基　　金：陕西省科技攻关计划(2022GY-092)。

摘　　要：针对物体表面残损区域存在明显的亮缺陷和不明显的暗缺陷这一特性,构建一种基于改进LOG算子与Otsu算法相结合的物体表面残损区域缺陷的边缘检测方法。首先,针对传统的LOG算子在检测图像亮缺陷边缘时检测结果不准确的问题,引入可根据图像特性自动调整模糊因子的Wiener滤波代替传统LOG算子中的高斯滤波,以提高图像亮缺陷检测的精度;其次,针对检测图像暗缺陷边缘时结果不准确的问题,使用Otsu算法分析图像暗缺陷的灰度直方图来自动确定阈值,以提升暗缺陷边缘检测准确率;最后,采用像素加权平均融合算法对检测出的物体表面图像亮、暗缺陷边缘进行融合,以实现物体表面残损缺陷检测。实验结果表明:相较于单独使用改进的LOG算子和Otsu算法,采用加权融合的方法检测到的缺陷像素点数量与原始图片中基本一致,能够更准确地对图像中物体表面残损区域的亮、暗缺陷边缘进行检测。In view of the characteristics that there are obvious bright defects and not obvious dark defects in the resid-ual area of the object surface,a defect edge detection method for the residual area of the object surface is constructed based on the combination of the improved LOG operator and Otsu algorithm.Firstly,aiming at the problem that the traditional LOG operator is inaccurate in detecting the bright defect edges of the image,Wiener filtering,which can automatically adjust the blur factor according to image characteristics,is introduced to replace the Gaussian filtering in the traditional LOG operator,thereby improving the accuracy of bright defect detection.Secondly,aiming at the problem that the dark defect edge detection result is inaccurate,Otsu algorithm is used to analyze the gray histogram of dark defects in the image to automatically determine the threshold,thereby improving the accuracy of dark defect edge detection.Finally,pixel weighted average fusion algorithm is used to fuse the bright and dark defect edges detected from the object surface image,so as to realize the detection of residual defects on the object surface.Experimental results show that compared with using only improved LOG operator or Otsu algorithm,the number of defective pixels detected u-sing the weighted fusion method is basically consistent with that in the original image,which enables more accurate detection of the bright and dark defect edges in the residual areas on the objects surface in the image.

关 键 词：表面图像残损检测 WIENER滤波 LOG算子 OTSU算法 图像融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]