Design And Implementation Of Foreground Obstacle Detection And Obstacle Avoidance System Of Mobile Robot

In recent years,the maturity of mobile robot technology has promoted the rapid development of related fields.As an important branch of indoor service mobile robot,it has begun to play a role in the fields of government affairs,education,medical care,logistics and other services.Detecting foreground obstacle and avoiding obstacle intelligently are an important research hotspot in indoor service-type mobile robots.In this paper,a Kinect sensor is used in the Linux system using C++ programming language and software libraries such as ROS,OpenCV library and libfreenect2 library is used to design a Mobile robot foreground obstacle detection and obstacle avoidance system.The dissertation first gives the overall framework design of the system,and briefly introduces the depth image restoration module,the foreground obstacle detection and obstacle avoidance module.Secondly,the depth image restoration module based on the tensor voting algorithm is designed to repair the missing information of the depth information in the depth image obtained from the Kinect sensor and improve the success rate and accuracy of the foreground obstacle detection module.Thirdly,using the output of the depth image restoration module,the background subtraction method is used to remove the interference of the background on the foreground obstacle.The initial obstacle detection is completed by image processing methods such as open computing,contour extraction,area filtering and convex hull processing.The minimum matrix constraint method is used to determine the number of obstacles and boundary information to provide valid input data for the obstacle avoidance module.Then,the obstacle avoidance module is calibrated by Kinect camera to convert the depth image coordinate of the obstacle into the world coordinate,and then completes the obstacle avoidance path selection based on the improved artificial potential field method.Because of the collision and the target unreachable,Local minimum point three questions,respectively,by correcting the gravitational function,the introduction of new repulsion function,the establishment of a temporary target point to solve.Finally,the paper tests the whole system and the main modules.The results show that the system has completed the preset target function and the system design has certain practical reference value.