Method and apparatus for standardization of multiple camera system

Abstract

The invention discloses a polyphaser system calibration method, comprising a calibration component which is formed by three or more signalized points with known geometry in rigid motion at the work area of the polyphaser system, obtaining a plurality of images of calibration and inputting the images to a computer, extracting image coordinate of the signalized points and regarding the image point which is shot by different cameras on the same signalized points on the calibration component at certain position as a group of corresponding points, linearly recovering the projection transformation matrix of every camera in the polyphaser system according to the geometry contained in the image corresponding points and the signalized points and utilizing nonlinearity arithmetic to optimize. The device disclosed by the invention comprises a calibration part, a handle, an operation system, a work area and a polyphaser system. The invention overcomes the shortages of low efficiency, having relatively large accumulated errors and the like in the existing calibration method which needs to calibrate the cameras in the polyphaser system one by one or step by step. Further, the invention realizes calibrating the whole polyphaser system at one-time, thereby having important practical engineering value.

Description

technical field [0001] The invention relates to the fields of computer vision three-dimensional reconstruction, motion capture and the like, in particular to the calibration of a multi-camera system. Background technique [0002] Multi-camera systems (including multi-camera or multi-camera systems) can be widely used in 3D reconstruction, visual monitoring, motion data capture in film and television, animation and game development and production, sports analysis and training guidance, as well as in medical research and rehabilitation medicine. Gait analysis and many other fields. An important step in using a multi-camera system is to calibrate the system, that is, to determine the projection matrix of each camera or to further decompose each projection matrix into internal and external parameters. [0003] The traditional multi-camera system calibration method generally uses a three-dimensional calibration block or a two-dimensional plane calibration template (Zhang Z Y. Fl...

AI Technical Summary

Problems solved by technology

However, Zhang’s method requires a point on the one-dimensional calibration component to be fixed, and the one-dimensional calibration component can only rotate around the fixed point. In the actual implementation, on the one hand, it is difficult to strictly rotate around the fixed point, which will seriously Affects the calibration accuracy; on the other hand, the operator's body will block some cameras, making it impossible to calibrate the entire system at the same time

This limits the practical application of the method

Images