Regional depth edge detection and binocular stereo matching-based three-dimensional reconstruction method

Abstract

The invention discloses a regional depth edge detection and binocular stereo matching-based three-dimensional reconstruction method, which is implemented by the following steps: (1) shooting a calibration plate image with a mark point at two proper angles by using two black and white cameras; (2) keeping the shooting angles constant and shooting two images of a shooting target object at the same time by using the same camera; (3) performing the epipolar line rectification of the two images of the target objects according to the nominal data of the camera; (4) searching the neighbor regions ofeach pixel of the two rectified images for a closed region depth edge and building a supporting window; (5) in the built window, computing a normalized cross-correlation coefficient of supported pixels and acquiring the matching price of a central pixel; (6) acquiring a parallax by using a confidence transmission optimization method having an acceleration updating system; (7) estimating an accurate parallax by a subpixel; and (8) computing the three-dimensional coordinates of an actual object point according to the matching relationship between the nominal data of the camera and the pixel andconsequently reconstructing the three-dimensional point cloud of the object and reducing the three-dimensional information of a target.

Description

technical field [0001] The invention belongs to the technical field of binocular stereo vision, and relates to the problem of processing real-shot images based on stereo matching, in particular to introducing edge detection technology into stereo matching, constructing support windows for pixels based on depth edges, and optimizing algorithms through confidence propagation A method for obtaining the matching relationship of pixels and reconstructing a three-dimensional point cloud model of a target object. Background technique [0002] Stereo matching-based binocular vision technology is currently a research hotspot in the field of 3D reconstruction. It restores the 3D information of the target by matching two images of the same target taken from different angles. Its advantages are mainly low requirements on hardware configuration, convenient and fast installation, and non-contact 3D reconstruction without any impact on the target. Moreover, through the 2.5D stereo dispari...

AI Technical Summary

Problems solved by technology

These matching algorithms contain some special matching strategies or manual steps, which affect the robustness and automation of stereo reconstruction; if the object changes, it may have a negative impact on the performance of the algorithm

In terms of reconstruction accuracy, since the main application of the binocular system adopts the stereo matching algorithm, the system is very dependent on the reliable disparity of pixels

The three main problems of stereo matching are: I) area occlusion II) low texture III) unbalanced lighting

However, for real images taken by binocular systems, these problems may be more challenging

Stereo matching in binocular reconstruction is a passive method based solely on the two images of the target without auxiliary information such as structured light. The reconstruction system is greatly affected by the shooting environment and the target itself

The actual environment is difficult to satisfy static lighting and balanced colors, resulting in the same scene easily forming different color information in the left and right images captured by the binocular camera, as well as unstable shadows and reflective areas

Moreover, the irregular shape change of the actual target is likely to cause complex occlusion areas and a wide range of possible parallax through the camera's shooting angle, further increasing the difficulty of matching

In the end, these problems in stereo matching lead to low quality of the 3D point cloud of the target object reconstructed by the binocular system

Images