Research On Mobile Robot Navigation Based On RGB-D Camera

Mobile robots have broad application prospects,and they have gradually penetrated into military,industrial,service,householdand other fields with the advancement of science and technology.However,the realization of autonomous navigation is the basic guarantee for mobile robots to perform various tasks.In this paper,a SLAM algorithm and obstacle sensing technology based on RGB-D camera of the mobile robots were studied,moreover,the obstacle avoidance strategy and global path planning method of the mobile robots were proposed.The main research works of this paper were summarized as follows:1.Aiming at problems like computing complexity and poor real-time performance of the current mobile robot SLAM algorithm,a real-time SLAM algorithm is proposed based on RGB-D camera.First,the FAST feature points in the RGB image are extracted,and then the direct method is used to minimize the photometric error to estimate the pose transformation of the camera.In order to reduce the accumulated errors in the motion of mobile robots,a closedloop detection method based on the bag-of-words model is proposed and the g2 o framework is adopted to optimize the pose graph.The experimental results show that the proposed algorithm can build a dense three-dimensional environment map with the real-time performance of SLAM system greatly improved(The experimental results show that the proposed algorithm can greatly improve the real-time performance of the SLAM system and build a dense threedimensional environment map).2.The data collected by the RGB-D camera is utilized to achieve the perception of obstacles in the environment.First,OTSU method is adopted to separate the background and the ground distractions are removed based on the geometric model.Then,the speckles in the image are removed by morphologic operations and all the convex hulls of the obstacles are acquired,thereby the position and size of the obstacles are obtained.In view of the dynamic obstacles,the CamShift algorithm is adopted to track the obstacles.Then the Kalman filter algorithm is performed to estimate the pose of the obstacle optimally and the obstacle velocity is estimated based on the optimized pose.3.Corresponding obstacle avoidance strategies are applied for different types of obstacles.The dynamic trajectory planning method is adopted to adjust the initial trajectory to achieve the avoidance of static obstacles,and the avoidance of dynamic obstacles is realized based on the idea of velocity vector.Finally,a physical experiment is conducted and the experimental results show that the mobile robots can safely avoid dynamic obstacles with a small cost.4.The global 3d path planning problem of mobile robots is studied under the condition of known working environment.First of all,the update strategy of ant colony algorithm global pheromone is improved.Then security factors are introduced,and an ant fallback mechanism under deadlock conditions is designed.A new search mode combining the octree search mode and the plane-by-plane search mode is proposed.The simulation results show that the length of the three-dimensional path obtained by the proposed method is short and the search efficiency is high,and it is more in line with the actual situation.The research results obtained in this paper have a good theoretical guidance and a certain practical value for improving the autonomous navigation capability of mobile robots.