基于DIC双圆交比的结构光标定方法  

Calibration method for structured light based on DIC double circle ratio

作　　者：礼冬雪 胡建军 吴迪 李墈婧 LI Dong-xue;HU Jian-jun;WU Di;LI Kan-jing(Institute 706,Second Academy of China Aerospace Science and Industry Corporation,Beijing 100854,China)

机构地区：[1]中国航天科工集团第二研究院七〇六所,北京100854

出　　处：《计算机工程与设计》2023年第5期1572-1578,共7页Computer Engineering and Design

摘　　要：针对结构光立体视觉的内部和外部参数计算中存在散焦态特征图案识别误差大的问题,提出基于DIC双圆交比的参数优化方法。为提取高精度的特征图案,设计面向结构光立体视觉参数优化的双圆特征图案。针对特征图案亚像素边缘提取,对基于Zernike矩的亚像素边缘提取算法进行优化研究,针对失焦产生的图像模糊问题,提出高精度、高效率的亚像素边缘检测算法。此外,利用双圆的交比不变性约束特性,获取采集图案的精确圆心,以及系统的初步标定参数。利用数字图像相关算法获取正视角下标定图案,圆心数据的线性约束获取圆心的精确位置,由此计算精确的系统标定参数。实验结果表明,该标定方法精度更高,稳定性更好。Aiming at the problem of large error in defocused feature pattern recognition in the calculation of internal and external parameters of structured light stereo vision,a parameter optimization method based on DIC double circle intersection ratio was proposed.To extract high precision feature patterns,a double circle feature pattern oriented to the optimization of structured light stereo vision parameters was designed.For subpixel edge detection of feature patterns,the subpixel edge detection algorithm based on Zernike moment was optimized.For the problem of image blur caused by out of focus,a high-precision and efficient subpixel edge detection algorithm was proposed.Using the cross ratio invariance constraint of double circles,the exact center of the collected pattern and the initial calibration parameters of the system were obtained.The digital image correlation algorithm was used to obtain the calibration pattern under the positive viewing angle,and the linear constraint of the center data was used to obtain the precise position of the center,so as to calculate the accurate system calibration parameters.Experimental results show that the calibration method has higher accuracy and better stability.

关 键 词：结构光 立体视觉 标定 交比特性 双圆靶标 边缘提取 校正算法 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]