The RGB-D SLAM Based On Kinect And Its Application Under Cloud Robot

Simultaneous Localization and Mapping(SLAM)in the unknown environment is an important research topic in the field of mobile robot navigation.With the development of RGB-D sensor technology,such as Kinect sensor,it has been favored by researchers in the field of robotics because of its high performance and low price.So,the research on RGB-D SLAM has also become one of the hottest research topics in the field of SLAM.Aiming at the shortcomings and shortcomings of RGB-D SLAM such as slow speed,complex computing task and huge data,this paper analyzes the front and back parts of the original RGB-D SLAM algorithm and improves several key links.Then based on the improved algorithm and combined RGB-D SLAM with the cloud robot and offload the back-end process of the RGB-D SLAM algorithm to the cloud in order to overcome problems such as large amount of calculation,low efficiency and low real-time performance existed in the traditional SLAM.The main detailed studying contents are as follows:First,we described each link of the RGB-D SLAM algorithm and analyze its shortcomings.According to these problems,this paper proposes specific improvement methods for three links including feature detection and extraction,motion transformation optimization and optimization of Pose Graph: for BRISK algorithm,using the pixel depth value for scaling instead of the Scale space Pyramid.The main direction of the feature points is determined by the Intensity Centroid;the optimization process of motion transformation is realized through combining ICP algorithm and particle filter;the HOG-Man algorithm is used as the core to optimize the Pose Graph.And then,we test the improved effect.Then,taking advantage of cloud computing in data computation and storage,we combine cloud robot with RGB-D SLAM algorithm,and offload the algorithm's back-end which has intensive computational tasks to the cloud.We presented a service framework based on the combination of these two aspects.Combined with the ROS system,the design method of ROS node is proposed,and the data storing and data sharing process between cloud and robot is given.Finally,the purpose which makes RGB-D SLAM more efficient and precise is realized.