Method for calibrating PTZ camera by using only two scene images

Abstract

The present invention discloses a method for calibrating a PTZ camera by using only two scene images. Firstly, two images with overlapped areas of any scene in a same focal length and different Pan-Tilt angles relative to the PTZ camera are photographed, the two images are subjected to feature point matching, a matching feature point pair between the two images is obtained, the influences of the uncertainty of matching feature point positioning and camera lens distortion are considered, a homography matrix estimation method with the combination of feature point positioning noise and lens distortion is created, the homography matrix between the two images and the radial distortion coefficient of the PTZ camera under the focal length are obtained, finally a Givens rotation matrix is constructed to decompose the obtained homography matrix, and thus the four internal references (focal length, principal point coordinate, and aspect ratio) of the camera is resolved and obtained. According to the method, the online calibration of five parameters of the PTZ camera can be realized.

Description

technical field [0001] The invention relates to a method for calibrating a PTZ camera with only two scene images, which can be applied to the calibrating of a PTZ camera in the fields of computer vision and video monitoring. Background technique [0002] PTZ camera (short for Pan, Tilt and Zoom) is a monitoring device that integrates high-resolution camera, variable-speed pan / tilt and optical zoom lens. Because of its high flexibility, it is widely used in various industries in modern industry. With the development of visual measurement technology, traditional static cameras with fixed parameters are not enough to meet the measurement work in some environments, such as large-scale measurement. The PTZ camera is used in the field of visual measurement, which solves the problems of small measurement viewing angle range and fixed viewing angle and direction, and reduces the requirements for the number of cameras while increasing the flexibility of the visual measurement system....

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that it requires a high-precision photography platform to achieve, and requires the movement information of the camera

But the disadvantage of this method is that it needs to use the information contained in the specific pattern in the scene to complete the calibration

Although this method overcomes the shortcomings of the above methods relying on scene information for calibration, its shortcoming is that at least three images need to be taken to complete self-calibration

It can be seen that the current general camera self-calibration method has two shortcomings of relying on scene information and needing to take multiple images. Therefore, this field needs a self-calibration method that can improve these two shortcomings at the same time, and it is expected that this method Can have higher precision and practical significance

Images