Calibrating method of hybrid vision system

Abstract

The invention relates to a calibrating method of a hybrid vision system. The calibrating method comprises the steps of: firstly respectively calibrating a monocular camera and a panoramic camera, obtaining intrinsic parameters of the monocular camera and the panoramic camera, arranging a chessboard diagram-based calibration plate, creating a chessboard coordinate system, then calculating a transformation matrix from a monocular camera coordinate system to the chessboard coordinate system, constructing a virtual view plane before a panoramic camera coordinate system, calculating a transformation matrix from the virtual view plane to the chessboard coordinate system, accordingly calculating a transformation matrix from the panoramic camera coordinate system to the chessboard coordinate system, and finally obtaining a transformation matrix from the monocular camera coordinate system to the panoramic camera coordinate system to obtain an external parameter of the hybrid vision system. The calibrating method is low in requirement of space arrangement relation of the calibrating condition and the hybrid vision system, flexible in use, and high in calibrating accuracy.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a hybrid vision system calibration method. Background technique [0002] In recent years, the hybrid vision system composed of monocular vision and panoramic vision has attracted the interest of many scholars. It combines the advantages of large resolution and clear imaging of monocular cameras with the advantages of large field of view and large amount of scene information obtained by panoramic cameras, and overcomes the shortcomings of small field of view of monocular cameras and low resolution of panoramic cameras. Prospects. In mobile robots, the panoramic camera is usually installed on the top of the robot and the monocular camera is installed at the front (supplementary viewing angle), which realizes functions such as object recognition, object tracking and three-dimensional reconstruction. Hybrid vision systems are also being used in the surveillance industry, wh...

AI Technical Summary

Problems solved by technology

The current calibration method has special requirements for the position of the calibration plate (such as perpendicular to the axis of the panoramic camera, and the calibration plate is placed in a known ideal position), which is difficult to guarantee in practical applications

[0006] 2. The optimization method used to solve the external parameters has certain limitations

As we all know, one of the common shortcomings of completely using the optimization method to solve the problem is that there are high requirements for the selection of the initial value. If the initial value is selected reasonably, the ideal result can be obtained, and if the selection is improper, the required value may not be obtained.

[0007] 3. The solution accuracy is low

From the existing method, it can be found that the deviation between the distance between the cameras and the actual value reaches 10mm, which increases the error of the reconstruction result

Images