基于M带小波变换的焊接熔池图像边缘检测  

Edge Detection of Welding Molten Pool Image Based on M-band Wavelet Transform

作　　者：薛家祥[1] 贾林[1] 李海宝[1] 

机构地区：[1]华南理工大学机械工程学院,广州510640

出　　处：《中国机械工程》2004年第13期1144-1146,共3页China Mechanical Engineering

基　　金：广东省自然科学基金资助项目 ( 0 0 0 3 80 )

摘　　要：利用M带小波变换理论对CO2 焊图像进行处理。基于Mallat小波模极大值边缘检测技术 ,研究了熔池边缘和焊缝位置的M带小波检测方法及算法实现 ,对边缘检测结果进行了分析。对CO2 焊图像处理结果表明 ,用M带小波变换理论构造的小波基对提取熔池边缘和焊缝位置效果较好 ,可用于熔深控制。The M-channel wavelet transform were employed to process the image of CO 2 welding in order to detect effectively the edge of the molten pool and the location of weld based on the approaches of Mallat's wavelet transform modulus maxima. The M-channel wavelet transform and the algorithm and the influences on the results of the edge detection have been investigated. It is demonstrated by the experimental results that better performance to extract the edge of the molten pool and the location of weld can be obtained by using the M-channel wavelet transform. The proposed edge detection approach can be further applied to the control of molten depth, the seam tracking and so on.

关 键 词：CO2焊 熔池 M带小波变换 边缘检测 

分 类 号：TG434[金属学及工艺—焊接]