Dual-camera calibration method and system

Abstract

The invention discloses a dual-camera calibration method and system. The method comprises the following steps: acquiring a plurality of groups of first images, second images and a movement matrix of a calibration plate; processing the plurality of groups of first images to obtain rotation matrixes and translation matrixes of the plurality of groups of first cameras; processing the plurality of groups of second images to obtain rotation matrixes and translation matrixes of the plurality of groups of second cameras; calculating a first translation matrix between the first camera and the second camera according to the rotation matrixes and the translation matrixes of the multiple groups of first cameras, the rotation matrixes and the translation matrixes of the multiple groups of second cameras and the multiple groups of movement matrixes, and determining a vector of the first translation matrix; determining a vector of the movement matrix according to the movement matrix; calculating a vector of a third translation matrix between the first camera and the second camera according to the vector of the first translation matrix and the vector of the movement matrix; and calculating the optical center distance between the first camera and the second camera according to the vector of the third translation matrix. The calibration precision is improved, and the calibration cost is reduced.

Description

technical field [0001] The invention relates to the technical field of machine vision, in particular to a dual-camera calibration method and system. Background technique [0002] Monocular vision systems and binocular vision systems are quite common in the field of machine vision. In terms of visual measurement, the monocular vision system mainly measures the size of the workpiece within the field of view of the camera. The measurement range and accuracy are restricted by the actual size of the workpiece, which cannot meet the needs of a large-scale visual measurement system. If a single camera needs to shoot high-resolution large The field of view requires a specially designed large field of view lens, which is too expensive and the system is not cost-effective. The binocular vision system can measure the size of large workpieces, and can measure the depth information, which can meet most of the visual measurement requirements. The traditional binocular vision system requ...

AI Technical Summary

Problems solved by technology

[0003] For the shortcomings of traditional detection methods of monocular vision and binocular vision, common solutions generally focus on improving the performance of cameras and lenses, but these solutions are costly and the effect is not significantly improved

Images