Research On Autonomous Navigation Technology Of Mobile Robot Based On Vision And Laser Information Fusion

With the development of social productivity and the improvement of science and technology level,mobile robot has been widely used in home,rescue,medical and industrial manufacturing.When there are many random obstacles,the space is closed and the light is dim,the mobile robot cannot achieve autonomous navigation by relying only on a single sensor.Therefore,this thesis studies the key technologies of Simultaneous Localization and Mapping(SLAM)and path planning involved in the autonomous navigation system of mobile robots.Based on the fusion of vision and lidar information,autonomous navigation in a complex environment is realized.Firstly,aiming at the problem of time and space unsynchronization of multi-sensor data,a corresponding synchronization algorithm is designed to unify the time stamps of different sensor data,complete the joint calibration of camera and lidar sensor,and obtain more accurate relative pose of the two.By comparing and analyzing the extraction and matching speed of different visual feature points,the optimal visual feature is selected for vision and Lidar fusion SLAM algorithm.Secondly,the SLAM scheme based on visual and laser information fusion is designed.When the robot localization fails or errors occur,the double insurance of visual loop detection and lidar proximity detection can effectively reduce the risk of robot localization failure.The fusion rules and Bayesian estimation method are adopted to realize the fusion and update of raster map constructed by camera and Lidar.Then,A* algorithm and Time Elastic Band(TEB)algorithm were selected as global and local path planning algorithms respectively,and the traditional A* global path planning algorithm was optimized.The optimized A* fusion TEB algorithm is used to interpolate the planned global path,set the temporary target points,and add constraints,and plan the route with high security and smooth curvature.Simulation analysis is carried out to verify its effectiveness.Finally,the mobile robot platform is built.By setting different real environments,the positioning and mapping effects,navigation accuracy and real-time obstacle avoidance effects of the mobile robot are verified.The experimental results show that the SLAM and the path planning algorithm can well meet the autonomous navigation requirements of mobile robots in real scenes,which further verifies the effectiveness of SLAM and the path planning algorithm.