摘要
针对轮式机器人平面运动的特性,将轮式机器人运动降到二维平面处理。由机器人相机的二维运动特征提出帧间特征点匹配等高约束条件,同时对帧间相机运动方程做线性化近似,提出直线约束条件。两个约束条件可用于误匹配特征点对点筛除,为视觉里程计帧间运动估计提供优质匹配点对。最后对筛选后的匹配点对进行降维处理,并与二维ICP算法相结合,用于估计机器人相机运动轨迹。试验表明,所提出的方法可大大提高视觉里程计运算速度。
According to the characteristics of the wheeled robot's planar motion,the wheeled robot's motion is reduced to a two-dimensional plane.Based on the two-dimensional motion characteristics of the robot camera,the inter-frame feature point matching equal height constraint conditions are proposed,and the linear approximation of the inter-frame camera motion equations is proposed to provide the straight line constraint conditions.The two constraint conditions can be used to filter out mismatched feature points,and provide high-quality matching point pairs for motion estimation between frames of visual odometer.Finally,the dimensional reduction of the matched point pairs after screening is combined with the two-dimensional ICP algorithm to estimate the motion trajectory of the robot camera.Experiments show that the method proposed in this paper can greatly increase the speed of visual odometer.
作者
李少波
王成功
任彦
巩泽辉
LI Shaobo;WANG Chenggong;REN Yan;GONG Zehui(School of Information Engineering,Inner Mongolia University of Science and Technology,Baotou 014010,China)
出处
《电工技术》
2021年第18期36-38,41,共4页
Electric Engineering
基金
国家自然科学基金(编号62063027)
内蒙古自然基金(编号2019MS06002)。
关键词
轮式机器人
特征点匹配
等高约束
直线约束
ICP算法
wheeled robot
feature point matching
contour constraint
straight line constraint
ICP algorithm
