摘要
针对环境亮度变化导致V-SLAM视觉里程计定位精度不准确的问题,提出一种基于改进ORB算法的视觉里程计定位方法。使用自适应阈值ORB算法提取特征点,提高特征提取的稳定性,通过FLANN进行粗匹配并采用PROSAC算法进行误匹配剔除,同时利用ICP方法进行图像配准求解位姿,使用光束法平差对轨迹图进行优化,采用TUM标准数据集和移动机器人验证算法的有效性。实验结果表明,该方法在大部分情况下定位精度优于其它算法,满足移动机器人的实际定位需求。
Aiming at the problem of inaccurate positioning accuracy of V-SLAM visual odometer caused by environmental brightness changes,a visual odometer positioning method based on improved ORB algorithm was proposed.The adaptive threshold ORB algorithm was used to extract feature points to improve the stability of feature extraction.FLANN was used for rough matching and the PROSAC algorithm was used for mismatch elimination.The ICP method was used for image registration to solve the pose,and the bundle adjustment was used to optimize the trajectory map,and the TUM standard data set and the mobile robot were used to verify the effectiveness of the algorithm.Experimental results show that the positioning accuracy of the proposed method is better than that of other algorithms in most cases,and it meets the actual positioning requirements of the mobile robot.
作者
张天宇
吴怀宇
陈志环
ZHANG Tian-yu;WU Huai-yu;CHEN Zhi-huan(Institute of Robotics and Intelligent Systems,Wuhan University of Science and Technology,Wuhan 430081,China)
出处
《计算机工程与设计》
北大核心
2022年第2期495-501,共7页
Computer Engineering and Design
基金
国家自然科学基金项目(61573263)
湖北省科技支撑基金项目(2015BAA018)
国家重点研发计划子课题基金项目(2017YFC0806503)。
关键词
视觉里程计
特征提取
自适应阈值
误匹配剔除
光速法平差
visual odometer
feature extraction
adaptive threshold
mismatch culling
bundle adjustment
