Design And Implementation Of Community Service Robot Navigation System

Robot navigation technology has always been one of the hot issues in the field of mobile robot.With the improvement of mechanization and the development of artificial intelligence technology,robots play an important role in human society.Community service robots work in modern smart communities and provide tremendous traversal for all aspects of people's lives.In order to ensure that the robot can adapt to a complex community environment and travel freely in the community,its navigation system must be powerful enough.This article through the research and analysis of existing service robot navigation technology,mainly completed the following work:1.Design and implement a complete community service robot navigation systems to single camera based,multi-sensor combination of hardware basis.The entire navigation system is developed based on the Robot Operating System(ROS),and different functional modules(such as sensor access module,road identification module,path planning module,etc.)are used as nodes to access ROS.This makes the entire navigation system has better maintainability and real-time.2.In the aspect of unstructured road recognition,this paper introduces the road recognition algorithm based on region classification and edge detection in detail.In order to extract the road area,The excellent Support Vector Machine(SVM)training classifier is used,and the HSI color features and a rotationally invariant local binary pattern(LBP)are used as image features.In order to extract the edge information of the image better,an improved Canny edge detection algorithm in color space is adopted.In order to extract the edge information of the image better,this paper adopts an improved Canny edge detection algorithm in color space,using neighborhood of 3 × 3,and calculating the gradient amplitude and angle in 4 directions.And in order to adapt to different environments,this paper uses the maximum between-class variance for each image to find the adaptive threshold,which makes the improved Canny edge detection algorithm more suitable for mobile robot autonomous navigation scene.Finally,the final results of the two algorithms are combined to get the final road image.3.In the aspect of obstacle avoidance,starting from a simple and effective template matching algorithm,its advantages and disadvantages are analyzed.Based on this,the existing researches are combined to optimize and improve it.First,the template image is rasterized,and then the obtained raster image is converted into a corresponding binary number.Finally,the template is matched by the bitwise AND operation of the binary number to find the optimal solution,which greatly improves the matching rate.Based on the above aspects,the subject has conducted multiple tests on the community service robot navigation system in the outdoor real environment.The test results have proved the integrity and effectiveness of the navigation system and have a certain theoretical significance and practical value for the research and development of community service robots.