Computer vision detection method for weld contour features based on discrete sequence points

Abstract

The invention discloses a computer vision detection method for weld contour features based on discrete sequence points. The method comprises the steps of 1 constructing a {XOZ} coordinate system according to the difference of the minimum resolutions of the sequence points in two mutually orthogonal directions; 2 constructing a bitmap capable of covering the whole discrete sequence points by taking the minimum resolution basic unit in the X direction as the sampling step length of the pixel points; 3 constructing a gray bitmap through coordinate values of the discrete sequence points in the {XOZ} coordinate system; 4 performing linear feature extraction on the constructed gray bitmap by using an LSD method in a computer vision algorithm, and obtaining related linear features at the upper edge and the lower edge of the contour; and 5 inversely calculating the real geometric information of the contour through the pixel coordinates of the obtained straight line to realize the quantitative description of the contour features of the weld. Compared with the prior art, the method has the positive effects that 1) the calculation efficiency is high; 2) the detection precision is controllable; and 3) the application range is wide.

Description

technical field [0001] The invention relates to a computer vision inspection method, in particular to a computer vision inspection method based on discrete sequence point description and oriented to weld contour features. Background technique [0002] Weld contour detection technology has a wide range of applications in welding engineering. The obtained contour information is the technical basis for application scenarios such as weld groove detection, automatic welding seam tracking, and welding assembly quality inspection, and is also an important embodiment of welding automation capabilities. Such as figure 1 As shown, wherein, the reference signs include: weld gap 1, weld step difference 2, weld centering 3, weld leakage groove 4, left base material 5, right base material 6; take butt weld as an example , the weld profile generally includes elements such as weld gap, weld step difference, and weld centering. The essence of weld profile detection is to abstract the geomet...

AI Technical Summary

Problems solved by technology

However, this type of method only uses the global statistical characteristics of the image to detect straight lines, and does not consider the overall information contained in the local discrete points in the picture, so it is easily affected by noise, resulting in false detection and missed detection of straight lines.

Images