基于RANSAC的相机标定优化算法  

Camera calibration optimization algorithm based on RANSAC

作　　者：姚广成 朱志峰 姚勇 YAO Guangcheng;ZHU Zhifeng;YAO Yong(School of Electrical and Information Engineering,Anhui University of Technology,Ma’anshan,Anhui 243002,China;Anhui FCAR Electronic Technology Co.,Ma’anshan,Anhui 243000,China)

机构地区：[1]安徽工业大学电气与信息工程学院,安徽马鞍山243002 [2]安徽省爱夫卡电子科技有限公司,安徽马鞍山243000

出　　处：《自动化与仪器仪表》2023年第4期6-10,共5页Automation & Instrumentation

基　　金：安徽省重点研究与开发计划:基于5G车联网的便携式机动车尾气监测及云诊断系统研发及产业化(2022107020012);深圳市科技创新攻关项目(JSGG20191129102008260)。

摘　　要：针对相机标定中外参标定精度低和标定速度慢的问题,设计一种X型标靶及其标定方法。对随机采样一致性算法(RANSAC)进行改进以及重新设计标靶,通过迭代更新得到最优单应性矩阵来优化相机外参旋转矩阵R和平移矩阵T。实验结果表明:该标靶减少算法迭代次数,提高了47%的标定速度,该RANSAC算法,新添应对存在多个最大值相同的方法,增加标定稳定性和精度;通过两组对比实验的分析,验证了该方法的可行性和稳定性,提高了相机标定速度。Aiming at the problems of low precision and slow speed of camera calibration with external parameters,an X type target and its calibration method are designed.The random sampling consistency algorithm(RANSAC)was improved and the target was redesigned,and the rotation matrix R and shift matrix T of camera external parameters were optimized by iterative updating to obtain the optimal homography matrix.The experimental results show that the target algorithm reduces the number of iterations and improves the calibration speed by 47%.The RANSAC algorithm adds a new method to improve the calibration stability and accuracy when there are multiple methods with the same maximum value.The feasibility and stability of the proposed method are verified by the analysis and design of two groups of comparative experiments,and the camera calibration speed is improved.

关 键 词：随机采样一致性算法 单应性矩阵 算法优化 实验分析 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]