Research On The Application Technology Of Robots Indoor Localization

With the continuous integration of information technology and industrialization,it has become an important symbol of modern technological innovation that robotics as the representative of intelligent industry is booming.Nowadays robotics technology is the hot topic in the field of engineering and technological research.China will incorporate robotics and intelligent manufacturing into priority areas for national scientific and technological innovation.Robots localization is the process that robots attains the information of itself and the surronundings by using various carried sensors,and calculates its pose after certain data processing.The robots indoor localization technology is the key point to the autonomous robots.The problem has been widespread concerned in the world of research institutions.According to different sensors,different robots indoor localizaiton technology have been proposed.In the paper,combined with the new technology of machine vision technology,the depth-sensor based method of robots indoor localization is proposed after analyzing the different localization technology.After analyzing the popular technology of robot simultaneous localization and mapping technology,and understanding the core idea of this method,the depth sensor is successfully applied to the robot indoor localization.Further work was carried out to deep understand the process of the depth image from mathematical principles.Pictures of different scale space are construncted with the color picture from the depth sensor.The image feature points are searched from different scale space and then matched between different images.The number of mismatch is reduced with the improve algorithm during the processing.Then the relative pose of the robot at different times is calculated according to the matched points between related images,as well as the corresponding depth information.In the calculation process,in order to improve noise robustness,RANSAC algorithtm is used for the initial estimates of relative position and orientation,and then ICP algorithtm is used for precise calculations.In the process of map expression,different location information is described by topological map.Topological map is optimized by the least squares method to obtain consistent global map.Hierarchical optimization algorithm is used in the optimization process in order to reduce the amount of calculation and achieve real-time optimization.During optimization,only the child nodes is optimizd from the parent nodes that are changed until the bottom of the original map.Real time optimization is achieved with less calculation.Other types of maps that we need are obtained by conversion of the optimized topological map.Related experiments were conducted to prove the effectiveness of the method.