Positioning method using double-positioning overall constraint

Abstract

The invention provides a method for improving the positioning accuracy and shortening the overall positioning time by using a double-positioning overall constraint method, shape models are extracted by using two relatively small shape features, after the shape models are matched, a vector is generated by the central points of the two shape models, an included angle of the two vectors is solved according to the cosine similarity of the two vectors, and Euclidean transformation is carried out on the ROI of the detection frame according to the included angle. The method can be applied to various visual detection systems, and solves the problems of very long overall positioning time, low detection efficiency, inaccurate positioning position and low detection accuracy when the to-be-detected image is very large.

Description

technical field [0001] The present invention mainly relates to the technical field of industrial machine vision image detection and pattern recognition, in particular, it is aimed at those image processing technologies that require fast and high-precision positioning. Background technique [0002] With the rapid development of image processing and pattern recognition technology, more and more industrial inspections have begun to use image processing algorithms to detect product defects, measure dimensions, target recognition, and target classification. [0003] For image processing algorithms such as defect detection of specific parts, target recognition in specific areas, and size measurement, if a large number of products are to be inspected cyclically, there will be an ROI positioning operation before each inspection starts, and the ROI is determined through the positioning function. position, and then perform image detection processing. [0004] Most of the existing loc...

AI Technical Summary

Problems solved by technology

[0006] In order to overcome the above-mentioned shortcomings of the prior art, the object of the present invention is to improve the positioning accuracy by using the overall constraint method of double positioning, and at the same time shorten the overall positioning time, which solves the problem of overall positioning when the image to be detected is large in the background technology. It takes a long time, the detection efficiency is low, the positioning position is not accurate, and the detection accuracy is low.

Images