Research Of Underwater Robot Welding Seam Image Recognizing Algorithms

With the rapid progress of exploring the sea,underwater welding robot plays an important role in the exploration of sea. So advanced techniques of image proccessing and image recognition are needed in the underwater welding robot based on vision system.Because of the speciality of imaging environment in the deep sea,the quality of image is poor,traditional image technique cannot meet the need of application.This paper has introduced the underwater welding seam identify system firstly,then studies of applying underwater image processing are carried out,the main contents are described as follows:1 An effective and robust image enhancement algorithm CLAHE (Contrast Limited Adaptive Histogram Equalization) is introduced for improving the contrast of underwater welding seam images. The CLAHE method gets a higher image contrast at a lower cost of visual appearance.2 A morphological operator of omni-directional multiple structuring elements is presented to filter the underwater welding seam image.This method can not only denosing but also can keep the features.3 After comparing with the 2D OTSU and C means clustering in the underwater welding seam image segment;Mean Shift can do segment very well and keep the features.4 Using morphological thin methold,Hough transform and projection to abstract the underwater welding seam feature,comparing Hough transform with projection methold,we choose Hough transform to abstract seam feature.