Multi-camera calibration method in large scene, information processing terminal and storage medium

Abstract

The invention belongs to the technical field of multi-view geometry under computer vision, and discloses a multi-camera calibration method in a large scene, an information processing terminal and a storage medium, and the method comprises the steps: carrying out the ambiguity matching filtering of multiple cameras in the large scene calibration through employing a scene clustering algorithm and an abnormal track removal algorithm; weak texture information enhancement is carried out on the ambiguous matched filtering image by using different fusion modes; and performing coordinate system registration on the image after weak texture information enhancement by using an ICP registration algorithm in combination with a phase mechanism type registration algorithm. According to the method, the adaptive capacity of the algorithm to texture similarity and texture sparseness in a large scene is improved. And the calibration precision is greatly improved. And high-precision multi-camera pose calibration in a large scene can be realized. According to the method, plane modeling and sphere center coordinate solving are carried out on the layout, and a conversion matrix is obtained. The above two modes complement each other, and the usability of the coordinate system registration algorithm is enhanced.

Description

technical field [0001] The invention belongs to the technical field of multi-view geometry under computer vision, and in particular relates to a multi-camera calibration method in a large scene, an information processing terminal and a storage medium. Background technique [0002] At present, one of the basic tasks of computer vision is to calculate the geometric information of objects in three-dimensional space from the image information obtained by the camera, and then reconstruct and identify the object, and the three-dimensional geometric position of a point on the surface of the space object and its corresponding point in the image The relationship between them is determined by the geometric model of the camera imaging, and these geometric model parameters are the camera parameters. Under most conditions, these parameters must be obtained by experiment and calculation. Whether in image measurement or machine vision applications, the calibration of camera parameters is ...

AI Technical Summary

Problems solved by technology

[0006] (1) In the existing technology, in the ambiguity matching and filtering of images from various perspectives, due to the ambiguity caused by similar textures, the calibration accuracy during matching is low

[0007] (2) In the enhancement of weak texture image information, the existing technology cannot avoid the occurrence of too dense feature points in the same area, so that the calibration accuracy under weak texture is very low

[0008] (3) In the prior art, in terms of image coordinate system registration, the existing enhanced coordinate system registration methods are poor in complementarity, which makes the registration image accuracy low

Images