摘要
针对平面运动轮式机器人的二维运动特点,对视觉里程计进行降维处理.使用汉明距与等高约束条件相结合,滤除误匹配特征点对与彩色图匹配正确但深度图误差较大的匹配特征点对.降维处理后使用二维ICP算法进行相机帧间运动估计.通过实验进行验证,可有效地提高平面运动轮式机器人视觉里程计的运算速度与计算精度.
According to the two-dimensional motion characteristics of plane wheeled robot,the visual odometry was reduced in dimension.The combination of Hamming distance and equal height constraints was used to filter out the matching feature point pairs of matching the color map correctly,but with larger depth map error.After the dimensionality reduction process,the two-dimensional ICP algorithm was used to estimate the motion between camera frames.It was verified by experiments that the operation speed and calculation accuracy of the plane motion wheeled robot visual odometry could be effectively improved.
作者
李少波
巩泽辉
崔桂梅
任彦
LI Shaobo;GONG Zehui;CUI Guimei;REN Yan(Information Engineering School,Inner Mongolia University of Science and Technology,Baotou 014010,China)
出处
《内蒙古科技大学学报》
CAS
2022年第3期280-284,共5页
Journal of Inner Mongolia University of Science and Technology
基金
国家自然科学基金资助项目(62063027)
内蒙古自治区自然科学基金资助项目(2019MS06002).
