Follow-up vision system optical axis and rotating shaft deviation calibration and external parameter correction method

Abstract

The invention relates to a follow-up vision system optical axis and rotating shaft deviation calibration and external parameter correction method. The method comprises the steps: establishing a mathematical expression of a camera optical center position Oc before rotation under a world coordinate system; defining a deviation parameter of the rotating shaft relative to the optical center of the camera; establishing a relationship between the optical center position of the camera before rotation and the optical center position of the camera after rotation; establishing a relative rotation matrixbetween the camera coordinate system before and after rotation; and obtaining a unit direction vector of the deviation parameter to obtain a coordinate vector corresponding to the rotation center O relative to the optical center of the camera, substituting the deviation parameter estimated value obtained by least square solution into a transformation relational expression, and executing correction of camera external parameters at any turntable rotation angle. According to the method, an inevitable shafting deviation model caused by manufacturing or mounting errors in a turntable vision systemis strictly derived, system errors in the process of correcting the external parameters of the camera according to the rotating angle are eliminated, and a one-dimensional model and analysis can be conveniently expanded to two dimensions and even three dimensions.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a method for calibrating the deviation between an optical axis and a rotating shaft of a servo vision system and correcting external parameters. Background technique [0002] The PTZ vision system is widely used in the fields of aerial photography, security monitoring and motion tracking, but its application potential in the visual measurement of large-scale spatial geometric quantities has not been fully developed. In order to obtain accurate three-dimensional measurement results in a unified reference system, the measurement results obtained by the camera at each rotation position are required to be transformed into a unified reference system through space transformation, which is essentially a process of camera extrinsic correction or attitude estimation. . [0003] Common camera pose estimation methods include the traditional calibration method based on cooperative t...

AI Technical Summary

Problems solved by technology

In practical applications that require arbitrarily rotating the camera, the above method cannot achieve online real-time extrinsic parameter correction

Even when the error of the turntable rotation angle is very small, due to the virtuality of the camera optical center and optical axis, the accuracy of the direct method of simply substituting the turntable rotation angle information to correct the camera extrinsic parameters cannot be guaranteed.

Images