基于PSO-OTSU算法的金属工件高光处理  

Highlight Processing of Metal Workpiece based on PSO-OTSU Algorithm

作　　者：费继友[1,2] 汪政 李花[2] 刘芝博 姜维 刘晓东[2] 张俊飞 FEI Jiyou;WANG Zheng;LI Hua;LIU Zhibo;JIANG Wei;LIU Xiaodong;ZHANG Junfei(School of Mechanical Engineering,Dalian Jiaotong University,Dalian 116028,China;School of Locomotive and Rolling Stock Engineering,Dalian Jiaotong University,Dalian 116028,China)

机构地区：[1]大连交通大学机械工程学院,辽宁大连116028 [2]大连交通大学机车车辆工程学院,辽宁大连116028

出　　处：《大连交通大学学报》2020年第3期28-33,共6页Journal of Dalian Jiaotong University

基　　金：国家科技支撑计划资助项目(2015BAF20B02);辽宁省教育厅科学研究计划资助项目(JDL2016026)。

摘　　要：针对在立体视觉测量系统中,因光照分布不均导致高反射率金属工件表面产生高光区域,进而影响视觉测量精度的问题,提出了一种基于PSO-OTSU算法的金属工件高光处理方法.以高反射率金属工件为实验对象,首先利用PSO-OTSU算法对高光区域进行了提取.通过与标准OTSU算法和GA-OTSU算法进行对比实验,该算法提取高光区域的结果更准确,提取时间分别缩短了78.88%和44.42%.其次利用直方图规定法对高光区域进行了处理,最后利用融合算法将处理后的高光区域与原始图像或非高光区域进行了融合,形成去高光图像.经实验证明,所提出的方法适用于多种金属工件场景图像的高光去除,且具有较好的去高光效果.In stereo vision measurement system,the surface of metal workpiece with high reflectivity is prone to generate highlight area due to uneven distribution of illumination,which affected the accuracy of vision measurement.To solve this problem,the Highlight processing of metal workpiece was proposed based on PSO-OTSU algorithm(Particle Swarm Optimization and OTSU algorithm).Taking the high-reflectivity metal workpiece as the experiment object,the PSO-OTSU algorithm was used to extract the highlight area.Compared with the standard OTSU algorithm and GA-OTSU algorithm(Genetic and OTSU Algorithm),the algorithm extractes the highlight area better,and the extraction time is respectively shortened by 78.88%and 44.42%.Then the histogram specification method was used to process the highlight region,and the fusion algorithm was used to fuse the processed highlight region with the original image or non-highlight region to form a de-highlight image.The final experiments show that the proposed method is suitable for highlight removal of a variety of metal workpiece scene images with good de-highlight effect.

关 键 词：金属工件 高光 PSO-OTSU算法 直方图规定法 融合算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TG174.4[自动化与计算机技术—计算机科学与技术]