A calibration method of stereo vision sensor based on spherical lens

Abstract

The invention discloses a stereoscopic vision sensor calibration method based on a spherical lens, which comprises the following steps: placing a checkerboard lattice target in a suitable position fora plurality of times, and photographing a target image refracted by a plurality of lenses; Extracting checkerboard grid corner points refracted by each lens in the image; The initial value of structural parameters is calibrated by analytical method. The optimal solution of structural parameters is obtained by nonlinear optimization. Compared with the perspective projection camera, the invention has larger field of view, and can realize three-dimensional reconstruction only by using one camera, and is suitable for large field of view measurement under narrow conditions.

Description

technical field [0001] The invention relates to sensor calibration technology and visual measurement technology, in particular to a spherical lens-based stereo vision sensor calibration method. Background technique [0002] Stereo vision sensors based on catadioptric elements have the advantages of compact structure, easy construction and low cost, and there is no synchronization error, which has attracted extensive attention of scholars in recent years. The stereoscopic vision sensor based on spherical lens is composed of a camera combined with two or more spherical lenses, which has a large field of view and is suitable for the measurement of a large field of view in a narrow space. [0003] The calibration of the stereo vision sensor based on the spherical lens includes two parts: the internal parameter calibration of the camera and the structural parameter calibration. There are many literatures about camera internal parameter calibration, so we focus on the structural ...

AI Technical Summary

Problems solved by technology

[0004] However, the existing main calibration methods based on projection array stereo vision sensors have relatively strict requirements on the relative position between the lens and the camera, such as requiring the camera to be parallel to the lens, which is difficult to achieve in actual operation; or requires manual intervention, For example, manually marking the outline of the spherical lens, the calibration process is complicated; or the accuracy is insufficient, especially at the edge of the spherical lens due to large aberrations and insufficient depth of field, resulting in poor imaging quality and inaccurate feature point extraction, resulting in poor calibration accuracy. Use the calibration results to complete the measurement

Images