针对成像性能不同的多个摄像机,提出了一种协同标定算法。算法主要由预处理数据和协同标定组成。预处理数据利用二分法快速找出错误特征点,并删除;协同标定使得一个精度高或观察点好的摄像机能帮助精度低或观察点不好的摄像机消除误差大的数据,提高标定精度,并对协同标定作了误差分析。实验表明,所提出的多个摄像协同标定算法适应性强,当数据组足够时可以明显消除外部实验环境给标定算法带来的误差。
A cooperating camera calibration algorithm based on cameras with different precision is proposed.This algorithm consists of data preprocess procedure and cooperating calibration procedure.Binary search algorithm is used in the data preprocess procedure to find the incorrect feature point,In the cooperating callbration procedure,the camera with high precision will help the low one(low-cost) to eliminate error data.The camera in good spot will help that in bad spot to eliminate error data,and the error is analyzed in detail.The experiments indicate that the cooperating calibration algorithm is robust,and the algorithm can eliminate the error data introduced by experimental environment when enough input data are collected.