一种基于点面特征的低漂移视觉里程计  

A low drift visual odometry based on point and plane

作　　者：金犇 贝绍轶[1] 尹明锋 李凌岩 JIN Ben;BEI Shaoyi;YIN Mingfeng;LI Lingyan(Jiangsu University of Technology, Changzhou 213000, China)

机构地区：[1]江苏理工学院,江苏常州213000

出　　处：《重庆理工大学学报（自然科学）》2022年第6期88-95,共8页Journal of Chongqing University of Technology：Natural Science

基　　金：国家自然科学基金项目(52172367);江苏省高等学校自然科学研究面上项目(20KJB520015);江苏中以产业技术研究院项目(JSIITRI202008);江苏省研究生实践创新项目(SJCX20_1041);常州市应用基础研究计划项目(CJ20200039)。

摘　　要：针对基于特征的视觉里程计在室内弱纹理环境下定位精度差等问题,提出一种基于点面特征的低漂移视觉里程计方法。首先,在提取ORB点特征和平面特征的基础上,构建联合重投影误差函数,并利用高斯牛顿法解算相机位姿。然后,计算曼哈顿主方向,利用环境的结构特性解决特征易丢失问题,降低相机运动过程旋转矩阵所积累的漂移误差。最后,在ICL-NUIM和TUM RGB-D数据集上进行评估。实验结果表明:所提算法在弱纹理、结构化环境下相比其他算法具有更好的精确性和鲁棒性。Aiming at the poor localization accuracy of feature-based visual odometry in indoor low-texture environment,a low drift visual odometry based on point and plane is proposed.Firstly,we construct a joint re-projection error function based on the extraction of ORB point features and plane features,and the Gauss-Newton method is used to obtain the camera pose.Then,the Manhattan main direction is calculated to solve the problem that features are easy to lose by using the structural characteristics of the environment,and the cumulative error of the rotation matrix is further reduced.Finally,the proposed algorithm is evaluated on ICL-NUIM and TUM RGB-D dataset.The experimental results show that the proposed algorithm has better accuracy and robustness than other algorithms in low-texture and structured environment.

关 键 词：视觉里程计 点面特征 弱纹理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]