Single-camera stereoscopic vision method

Abstract

The invention belongs to the technical field of 3D sensor measurement, and provides a single-camera stereoscopic vision method, which comprises the following three operation steps of: 1, image acquisition: acquiring N (N is greater than or equal to 2) spectral images by utilizing a snapshot type multispectral camera device when one-time exposure is carried out; 2, focusing definition calculation:the image processing and analyzing device calculates the focusing definition of the N more than or equal to 2 spectral images at each pixel (X, Y spatial positions); and a third step of depth detection: according to the focusing definition, obtaining longitudinal (i.e., depth) Z information of a position corresponding to each pixel, and completing XYZ stereo vision. The method overcomes the defects that binocular stereoscopic vision computation is large and a laser 3D vision system is high in price, and has wide application value in advanced manufacturing, intelligent robot movement, unmannedautomobile navigation, obstacle avoidance and the like.

Description

technical field [0001] The invention relates to machine vision, in particular to a single-camera stereo vision method. Background technique [0002] The natural world is three-dimensional, with plane (X, Y) position information and Z depth information. There are three main methods for people to recognize depth information: through the parallax information of the left and right eyes, through the motion information of the eyes relative to the movement of the object, and through the surface color information of the object. There are also those that detect depth through object surface texture information. [0003] In the past 20 to 30 years, the 3D vision methods studied by the academic circle mainly include obtaining depth information by analyzing the disparity information of the left and right eyes. Although 3D vision has been discussed for a long time in the literature, most machine vision systems today often rely on the analysis of two two-dimensional images acquired by tw...

AI Technical Summary

Problems solved by technology

[0005] The main purpose of the present invention is to overcome the disadvantages of the existing machine vision methods, especially the dual-camera stereo vision method, that the calculation amount is large and the real-time performance is poor. The existing 3D laser vision system is expensive and the detection results are sparse clouds without realistic 2D image information. disadvantages, a single-camera stereo vision method is proposed