Research On Mobile Robot's V-SLAM And Indoor Navigation Using Kinect

At the beginning of the creation of computer vision,people looked forward to the day that computers could look through the eyes to the world,understand the surrounding things,explore the unknown environments,and complete the tasks of navigation.With the rapid development of artificial intelligence and machine learning technology,mobile robot SLAM and indoor navigation have become the focus of research in the field of robotics research.Therefore,it is of great theoretical significance and practical application value to study the visual SLAM and indoor navigation research of mobile robot based on Kinect.The main work of this article is as follows:Firstly,research status,navigation technology and common sensors of mobile robots are studied.The robot operating system is selected as the system software platform,and the Kinect camera is used as the environmental information acquisition device.The visual SLAM and indoor navigation scheme based on Kinect is completed.Secondly,for the problem that the iterative nearest point algorithm is easy to fall into local optimum and the matching error is large,a new double restriction method include Euclidean distance and Angle threshold is proposed to complete the exact matching of the point cloud;in order to eliminate the accumulation in the visual SLAM Errors,improved similarity score function in loop detection,reduced perceived ambiguity,improved closed loop recognition rate;Weights are introduced into the key frame selection,and the pose of the robot is optimized based on the g2o(general graph optimization)algorithm.Based on this,the visual SLAM system of the indoor mobile robot based on Kinect is constructed.Experimental results show that the improved algorithm is robust and can effectively improve the accuracy of map construction and reduce system errors.Thirdly,study the artificial fish swarm algorithm.For the traditional AFSA,the random selection of the field of view and the step size will make the optimization of the algorithm less accurate.An improved Artificial Fish Swarm Algorithm is proposed.Firstly,reverse learning is introduced to optimize the distribution of artificial fish in the problem space and obtain a better fitness value than the initial position.Then introduce improved adaptive factors to automatically adjust Visual and Step.Inspired by the PSO algorithm,the global and local search capabilities of the inertia weight factor balance algorithm are finally introduced.The experimental results show that the improved algorithm can effectively improve the convergence speed and accuracy of the algorithm and plan the optimal path.Finally,the visual robot SLAM and indoor navigation system based on Kinect was designed and implemented on the Pioneer-3DX mobile robot platform.The robot vision SLAM and autonomous navigation experiments were completed in the indoor environment.Experimental results show that the visual SLAM based on Kinect for mobile robots presented in this paper is highly adaptable to indoor navigation systems and has high navigation efficiency in complex environments.It has high feasibility and reliability.