Processed RGB-D SLAM based on HOG-Man algorithm. (3rd September 2019)

• Record Type: Journal Article
• Authors: Liu, Yanli; Zhu, Mengyu
• Is Part Of: International journal of high performance computing and networking. Volume 14:Number 3(2019)
• Page Start: 376
• Page End: 384
• Publication Date: 2019-09-03
• Subjects: RGB-D SLAM -- mobile robot -- HOG-Man algorithm -- Kinect; High performance computing -- Periodicals; Computer networks -- Periodicals; High performance computing; Periodicals; 004.05
• Languages: English
• ISSNs: 1740-0562

Abstract

Simultaneous localisation and mapping (SLAM) of a robot is the key to achieve autonomous control of the robot and also a significant topic in the field of mobile robotics. Aiming at 3D modelling of indoor complex environment, this paper presents a fast three-dimensional simultaneous location and mapping method for mobile robots. On the basis of the HOG-Man algorithm which is the core of RGB-D SLAM algorithm, the open-source software combining the RGB-D sensor like Kinect with the wheeled mobile robot is used to obtain the odometry data, and then the information of the location is matched through the image feature extraction and in the end the map is optimised by HOG-Man algorithm. Finally, the feasibility and effectiveness of the proposed method are verified by experiments in indoor environment.