Research On The Method Of Mirror Detection Based On SLAM In Indoor Environment

The research on taking robot as the carrier has become the focus of scholars from all walks of life.Simultaneous localization and mapping(SLAM)is one of the key technologies in the field of robot.Laser radar with its high precision as perception sensor is widely used.In the process of scanning,when the environment includes a mirror,distance of the mirror can not be measured by laser radar.So it will not provide the robot with accurate environment information,which causes distress on obstacle avoidance and path planning of the robot.It will be very meaningful to solve the issue.In this paper,first,the SLAM method based on the Rao-Blackwellized particle filter(RBPF)is used.Laser radar and odometer are used as the sensing sensor,and the environment localization and map construction of the mobile robot are realized.RBPF has been widely used to solve the SLAM problem of robots.It can achieve accurate positioning of robots and map construction,and it reduces the computational complexity.Secondly,in order to solve the problem of laser radar scanning which doesn't attain the distance of mirror,a novel method is proposed for detecting mirror information.In this paper,according to mirror imaging will move with the movement of the robot,the position of mirror will be judged from the symmetry relationship between the position of the robot and the mirror image And probability model is used to determine the existence of mirror in the moving process.Finally,the robot's positioning and construction of the environment are realized in different experiment environments.And the validity and applicability of the proposed mirror detection algorithm are verified according to multiple experiment results.