基于数学形态学的水下焊接图像处理  

UNDERWATER WELDING IMAGE PROCESSING BASED ON MATHEMATICAL MORPHOLOGY

作　　者：张为民[1] 钟碧良[1] 

机构地区：[1]广州航海高等专科学校计算机与信息工程系,广东广州510725

出　　处：《计算机应用与软件》2011年第5期131-134,共4页Computer Applications and Software

基　　金：广东省高等学校自然科学研究重点项目(06Z028)

摘　　要：水下拍摄的激光结构光焊接图像不可避免要受弧光、飞溅以及烟尘的强烈干扰。根据水下焊接图像特点,采用图像加窗、锐化、阈值分割、滤波以及边缘检测等步骤来完成针对水下焊接熔宽的检测。通过各种方法的比较,说明数学形态学更适合水下焊接图像处理。Welding image captured underwater with laser structured light is inevitably interfered strongly by arc light,splash and soot.According to the characteristic of underwater welding image,in this paper the detection aiming at weld width of underwater welding is implemented through the procedures of window processing,sharpening,threshold segmenting,filtering as well as edge detecting,etc.The mathematical morphology method is shown more suitable in processing underwater welding image after comparing various methods.

关 键 词：水下焊接 图像处理 数学形态学 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]