基于图像分析的摄像头轴偏量标定方法

Calibration of camera axis offset based on image analysis

针对摄像头在焦距调节过程中,视频图像的中心点在物体上滑移的问题,进行了理论建模分析,将传统的分段测量、曲线拟合的解决方法进行了改进,提出了一种基于图像分析的轴偏量的标定方法,利用矩形图像确定不同焦距下摄像头的轴偏量,并在实验验证中进行了优化处理。结果表明,本方法可以较为精确地标定出摄像头轴偏量,经过修正补偿后的摄像头,轴偏抖动可以控制在视场角的0.6%,可以满足较高精度闭环跟踪场合的要求。

For the problem that objects in image center is excursing while camera is zooming,this paper analyses the modeling and improves the traditional method.Based on the results,a calibration method of camera axis offset based on image analysis is put forward.Moreover,some optimizations are processed in the validation experiment by rectangular object imaging with different focal length,whose results show that the method could accurately calibrate the camera axis offset.After calibration of camera,axis deviation jitter is lower than 0.6% of view angle,which could provide facility for using camera under high precision situations.