Image processing algorithm of weld seam based on crawling robot By Binocular Vision

Abstract

By Binocular Vision way, this paper aimed to research welding image process of weld seam based on crawling robot. It built a developed suitable tracking control policy and software flatform. Some filters have been studied, such as mean filter, median filter, homomorphic filter. A cascade algorithm composed by medium value filter and homomorphic filter was used to welding image de-noising, which had been proved effective. Besides, an adaptive threshold arithmetic for segmented image was put forward, which had a good robustness, namely cascade OTSU(Maximum Classes Square Error). Al-alloy butter plate welding experiment proved that welding torch could track seam well.