Research Of Mobile Robot’s SLAM Based On Binocular Vision

Mobile robots technology is one of the most important subjects in scientific research,and the robot’s binocular vision system in robots navigation is of great conclusiveness.What the robots concerned most in practical application are “Where am I?” “Where am I going?”and“How to get there?”.To solve these three questions at one time,the Simultaneous Localization and Mapping(SLAM)is being presented.Firstly,this topic did some analyses of SLAM’s research,illustrated some development issues of SLAM.To finished the robot’s route planning by using the binocular vision robot’s two cameras,firstly we should extract the feature points and get the matching points.On this experiment we dealt with the image processing issue by using SURF arithmetic,and then used new matching points as new landmarks to finish the robot’s route planning.On this experiment,we use the tracked robot to discuss the SLAM’s problem.The tracked robot moving smoothly,so the pictures it captured had less errors.In that case,the tracked robot could meet the demands of SLAM ’s research in binocular vision system.This topic also illustrated a modified EKF-SLAM based on the neural network support to improve the feature map’s accuracy.By doing experiment,we concluded that the m NNEKF-SLAM is better than EKF-SLAM on feature map’s building.Mobile robot’s SLAM research is based on binocular vision system’s cameras.And the camera’s another application is in video surveillance.As with the developing of modern society,intelligent monitoring is becoming more and more urgent.This topic also presented an arithmetic based on video surveillance,which could improve the efficiency on video surveillance by saving a lot of time.