Research Of Quality Evaluation Of Weld Formation Based On Stereo Vision

Welding technology is widely used in the field of industry.Automatic welding based on vision and image processing technology is also a hot issue in the current research.However,the main way to evaluate the quality of traditional weld forming is manual measurement or contact measurement with tools.The efficiency is very low,and the data fluctuate easily.This paper introduces the computer vision and using binocular stereo camera to evaluate the weld forming quality and improve the traditional measurement way,which has strong practical significance.In this paper,binocular camera is used as a visual sensor to collect weld images.Through coordinate transformation and camera calibration,we get the internal and external parameters of the camera.Then we use Bouguet binocular stereo correction algorithm to calibrate the camera to achieve the camera alignment.The weld image is affected by the environment and the metal welding board,and it is easy to produce a lot of noise.In this paper,the median filter algorithm is used to denoise the weld image.Considering that the traditional two-value segmentation algorithm can not accurately extract the weld area problem,this paper proposes a grab-cut based image two value segmentation algorithm,which can extract the weld area well.After extracting the contour of the weld,it is necessary to calculate the 3D coordinates of the weld edge to calculate the parameters of the weld forming quality.In this paper,the edge detection algorithm of Canny is used to extract the edge of the weld area.In the calculation of space information of binocular shooting objects,the traditional method generally adopts dense stereo matching method,and there is a lack of parallax and mismatch.In this paper,the principle of triangulation is used to calculate the space coordinates of the weld.It can effectively use the characteristics of the binocular camera to accurately obtain the height information of the weld.On the basis of theoretical research,this paper constructs an experimental system for evaluating the quality of weld formation based on stereoscopic vision,and carries out the experimental test The binocular camera is used to collect weld images,denoise the edges of the binocular weld and extract the outline edge,and draw the system software interface with MFC.According to the existing standards,the evaluation index of weld quality is worked out,and the quality of weld formation is evaluated.The experimental results show that the weld forming precision of this paper can reach about 0.5mm,and the quality of weld forming can be evaluated well.Finally,the research work is summarized,and the future research content and direction are prospected.