Method for straight line stereo matching

Abstract

The invention provides a method for straight line stereo matching, which can build a model aiming at problems of screening and uncertainty about characteristics in straight line stereo matching and realizes solving of all kinds of congruent relationships including 'one to one', 'one to many' or 'many to many' among straight lines. The workflow for solving the problem in the invention is as follows: reverting corresponding projection straight line from statistical significance by using extracted straight line characteristics, generating all the straight line characteristic groups from reverted projection straight lines, and selecting the most reliable straight line group as a solution to all the matching problems. The way of estimating end points of the projection straight line by searching effective edge points can improve the reliability of the estimation of the end points of the projection straight line segment, facilitate the realization of characteristic grouping of the reverted projection line segment by adopting the method of solving maximum clique in graph theory and enhance the accuracy of the matching result by counting the reliability of every straight line characteristic group.

Description

technical field [0001] The invention belongs to the technical field of photogrammetry and image processing, and in particular relates to a straight line matching method. Background technique [0002] Stereo matching is a long-term researched fundamental problem in the field of photogrammetry and image processing. Its purpose is to obtain the 3D information of the target scene by matching feature sets extracted from two images from different viewpoints. Straight lines are an important feature of stereo matching applications, and many scholars at home and abroad have devoted themselves to the problem of stereo matching based on straight lines. [0003] At present, straight line stereo matching is a recognized problem, which is mainly reflected in the following two aspects: First, due to the influence of scene occlusion, image noise and uncertainty in the process of line extraction, resulting in the same name extracted from different images The attributes of straight line feat...

AI Technical Summary

Problems solved by technology

The disadvantage of these two grouping methods is that the straight lines are grouped in different images. Due to problems such as projection transformation and occlusion, the grouping results may change from one image to another, and it is difficult to ensure the stability of the grouping process, and , except for the compatibility of a few broken straight lines, they cannot completely solve all the correspondence problems of "one matches many" and "many matches many", especially complex correspondence problems involving multiple non-collinear straight lines

Images