An anti-interference sub-pixel line fitting method

Abstract

An anti-interference sub-pixel line fitting method, which sets the ROI of the image, detects the edge points of the line in the ROI, and then uses the neighborhood weighting method to perform sub-pixel refinement on the edge points, and calculates the sub-pixel gradient of the edge points Value, through the comprehensive judgment of the gradient threshold and polarity direction, the points that meet the conditions are subjected to quadratic curve fitting and smoothing filtering, and recorded as candidate straight line edge points; then use the straight line fitting algorithm based on the distance weight of the edge points to The edge points of the straight line are fitted iteratively for multiple times to obtain the parameters of the straight line, and the intersection of the fitted straight line and the ROI is calculated to obtain the two end points of the final target straight line. The present invention proposes a high-precision straight line fitting algorithm based on edge point distance weights, using multiple iterative fitting strategies, which greatly reduces the interference of outliers and noise points on the target straight line fitting, and has high The advantages of stickiness, high precision and strong anti-interference ability.

Description

technical field [0001] The invention belongs to the technical field of machine vision and image processing detection, relates to sub-pixel line detection, and is an anti-interference sub-pixel line fitting method. Background technique [0002] With the development of industry, the size requirements of workpiece materials are getting higher and higher. Accurate detection of workpiece size and other information plays an important role in precision electronics manufacturing and precision machinery manufacturing. At present, most of the linear edge detection of workpieces in China is only at the pixel level, and due to the influence of workpiece processing factors, the anti-interference ability of the detection results is poor. Pixel-level, unstable line detection can no longer meet the needs of high-precision industrial manufacturing, which limits the development of high-precision industries. [0003] The traditional line detection technology usually uses edge detection operat...

AI Technical Summary

Problems solved by technology

[0006] The problem to be solved by the present invention is: In order to solve the problems of low precision of traditional pixel-level line detection and poor anti-interference ability, the present invention proposes an anti-interference sub-pixel line fitting method; the present invention effectively solves the problem of pixel-level edge The problem of insufficient detection accuracy can effectively suppress the interference of outliers on the edge of the line, and finally achieve the effect of high-precision and anti-interference detection of line information

Images