Scene depth restoring and three dimension re-setting method for stereo visual system

Abstract

The invention relates to a scene depth recovery and three-dimensional reconstruction method of a stereo vision system, which obtains internal parameters, relative rotation matrices and translation vectors of left and right cameras through calibration; matches corresponding points in left and right images; for a pair of matching points, Calculate their left and right normalized vectors from their left and right image coordinates and the corresponding camera internal parameters; according to the coordinates of the left and right normalized vectors, take the depth of the scene point in the left and right camera coordinate system as a parameter, and obtain the light from the left camera respectively Two projection line equations from the center to the left image point and from the optical center of the right camera to the right image point; from the left and right normalization vectors and the relative rotation matrix and translation vector of the camera, the common perpendicular endpoints of the two projection lines are obtained in the left and right cameras Depth value and three-dimensional coordinates in the coordinate system; the coordinates of the midpoint of the common perpendicular are obtained from the coordinates of the two endpoints of the common perpendicular, and the coordinates of the midpoint are used as the three-dimensional coordinates of the scene point to complete depth restoration and three-dimensional reconstruction.

Description

technical field [0001] The invention relates to a scene depth recovery and three-dimensional reconstruction method of a stereo vision system, in particular to a direct method for scene depth recovery and three-dimensional reconstruction from a stereo vision system with any configuration, which is suitable for scenes in computer vision and photogrammetry Depth restoration and 3D reconstruction. It belongs to the field of advanced manufacturing and automation. Background technique [0002] Stereo vision has been an important problem in computer vision and photogrammetry. A stereo vision system generally includes two or more cameras, and the internal parameters and relative positional relationship of these cameras become known data through calibration. The so-called scene depth recovery is to calculate the distance of a scene point relative to a certain camera along the main optical axis according to the corresponding image point coordinates of a scene point in two or more ca...

AI Technical Summary

Problems solved by technology

The second type of method usually requires initial value estimation first, and then iterative optimization, which is time-consuming and easy to converge to a local optimal solution, so it is rarely used in practical applications.

Therefore, the first type of method is generally used in stereo vision at present, but when using the first type of method, image correction needs to be performed first, which is also a time-consuming step, and errors will be generated during the transformation process of image correction, which will affect the reconstruction accuracy

Images