Research On Mobile Robot Indoor Positioning System Based On Ground Feature

As large number of robots are used in industrial manufacturing and life services,robots have assumed an increasingly important role in human society.For robots to work accurately,positioning is a very important part.The robot can be positioned by a Global Navigation Satellite System(GNSS)in the outdoor environment.However,the positioning accuracy of GNSS indoors often fails to meet the requirements.Therefore,the positioning of the robot in the indoor environment has become a hot spo t for domestic and foreign research.In recent years,due to the continuous reduction of the cost of visual sensors,and compared with indoor positioning technologies such as WIFI,infrared and ultrasonic waves,it has obvious advantages in positioning acc uracy and processing speed,making the indoor positioning of robots based on visual sensors widely attention.Therefore,this thesis proposes a combined positioning method that combines visual positioning and odometer track estimation to achieve real-time positioning of mobile robots.The main research contents and innovations of the thesis are as follows:1.By referring to the mainstream indoor positioning technologies of robots at in our country and abroad,the limitations,advantages and disadvantages of different positioning methods are summarized,the indoor location characteristics of robots used in life are analyzed,and a mobile robot indoor positioning system based on ground characteristics is proposed.2.According to the robot motion and the positioning method adopted in this dissertation,a mobile robot hardware platform equipped with a wide-angle camera at the bottom is built,and carry out algorithm research and experimental test on this mobile robot platform.3.Calibrate the camera installed at the bottom of the mobile robot platform,get the internal and external parameters of the camera and the lens distortion coefficient;then perform Canny edge detection and 8 neighborhood edge grouping on the collected grayscale image,and then perform image distortion correction on the edge image,through a comparative experiment proves that the image preprocessing has higher operation efficiency.Aiming at the characteristics of the ground square straight lines in the image,the advantages and disadvantages of mainstream straight-line fitting algorithms are analyzed and compared.A straight-line fitting algorithm combining pre-tested Random Sample Consensus(RANSAC)algorithm and least square method is proposed,and the effectiveness of the algorithm is verified by comparative experiments.Finally,the Inverse Perspective Mapping(IPM)algorithm is used to convert the fitted ground square straight line to the world coordinate system,and calculate the posture data of the straight line in the camera coordinate system,as the output information of the vision module.4.The odometry calculation model of the mobile robot odometer is establish ed.By analyzing the principle of the Kalman filter algorithm,the ext ended Kalman filter algorithm is used to fuse the information of the mobile robot's position sensor and visual sensor.Based on the ground characteristics of the indoor environment,a pose correction based on line prediction algorithm is proposed.Experimental results show that the system can achieve high-precision positioning,can effectively eliminate the influence of ground interference lines,and has good innovation and engineering application prospects.