Research Of Mobile Robot Simultaneous Localization And Mapping Based On Multi-sensor Fusion

In order to achieve higher autonomy and adaptability,mobile robots must be able to achieve autonomous navigation in unknown environment.The key technology to achieve autonomous navigation is the Simultaneous Localization And Mapping(SLAM)of mobile robot.Which is,the robot can move according to its own sensors while moving in an unknown environment,and continue to create incremental maps while moving.In the current SLAM research,it is mainly divided into laser SLAM and visual SLAM according to the different types of sensors used by robots.Laser SLAM uses a laser as a sensor of a robot.It has high accuracy of construction and good real-time performance,However,it can obtain a small amount of information,a large cumulative error,and can only detect obstacles on the installation plane.Vision SLAM uses a camera as a sensor.It can obtain rich environmental information,but its mapping accuracy is low.Due to the large amount of information which needs to be processed,the Vision SLAM real-time performance is poor and it is easily affected by the environment and light.Therefore,using a single sensor to conduct SLAM research has various defects,and it is difficult to create a high-precision and complete information map in practical applications.Aiming at the various defects of the above single sensor,A method is presented in this paper to fuse these two sensors and use the fusion data to make an environment map.This method first converts the depth map collected by Kinect into pseudo laser data,and then merges it with the information collected by the laser sensor.After that,using the fused information to construct a local grid map of the environment.Then uses the fused data to make a laser loop closure detection.When a loop is detected,use Kinect's RGB image for visual loop detection.The double loop closure detection is used to filter out the wrong loop closure,which not only improves the accuracy and completeness of the global map,but also avoids the problems of poor real-time caused by a lot of visual information mapping.Finally,in the experimental environment built by ourselves,the robot operating system(ROS,Robot Operating System)is used to conduct experiments and tests the different sensors to verify the feasibility of the program.The experimental results show that the environment map created by the fusion of the two sensors has a greater advantage in accuracy and integrity than the single sensor...