Aiming at RT weld image,a method of edge detection which combines sobel operator and dynamic adaptive threshold is put forward.

Recognition of welding image based on improved C-V method;

All-fiber optic integration type of weld seam image sensing system

This article carries out feature analysis and summary of V sharp weld seam image collected by structural-light,and proposals different image processing methods aiming at different feature,on basis of which,weld seam image process is completed by means of global threshold segmentation and morphology filter determined by automatic threshold,and obtain excellent effect of image processing.

Defect extraction in weld image based on Generalized Fuzzy Operator;

Some filter methods were analyzed to find the characteristics of the weld image noises.

Improved OTSU method on welding seam image segmentation;

However, because welding seam images are complicated, it is very important to choose suitable and effective processing algorithm in order to get accurately the information.

The wavelet edge detecting method was introduced to the welding seam image processing in this paper,which can make up the defect of usual edge detecting methods in antinoise ability and precise locating ability.