基于免疫遗传小波分析的管道焊缝图像处理方法  

Pipeline weld image processing method based on immune genetic algorithm and wavelet analysis

作　　者：徐平[1] 张方舟[1] 常洪庆[2] 严胡勇[1] 巩淼[1] 姚姜虹[1] 

机构地区：[1]东北石油大学计算机与信息技术学院,黑龙江大庆163318 [2]大庆油田热电厂除尘分厂,黑龙江大庆163314

出　　处：《东北石油大学学报》2013年第1期110-115,11-12,共6页Journal of Northeast Petroleum University

基　　金：黑龙江省教育厅科学技术研究项目(11551016)

摘　　要：针对难以实现管道焊缝图像信息准确提取的问题,利用免疫遗传算法优化小波系数,将小波变换与免疫遗传算法引入焊缝图像的处理过程,采用相似性距离分析传统小波、Canny算法、免疫遗传小波等算法优劣性.结果表明:该算法能够有效地抑制噪声干扰,去除图像边缘中的伪影,从而较好地实现图像边缘信息的清晰与细化,体现遗传小波变换算法在精确定位与抗噪声能力上的优越性.It is difficult to realize the accurate extraction of weld image information in the petroleum pipeline at present.In view of the problem,this paper introduces wavelet transform and immune genetic algorithm to the treating process of weld image according to the immune genetic algorithm to optimize wavelet coefficient,meanwhile it uses similarity distance to quantitative analysis traditional wavelet,Canny algorithm and immune genetic algorithm's quality,according to the calculation and comparison of similarity distance,we can draw a conclusion:this algorithm can realize the clarity and refining of image border better and reflect the superiority of genetic wavelet transform algorithm's ability in the accurate positioning and anti-noise adequately.

关 键 词：免疫遗传算法 小波分析 焊缝 图像处理 相似性距离 管道 

分 类 号：TG441.7[金属学及工艺—焊接]