Title :
Visual odometry - A review of approaches
Author :
Boxin Zhao;Tianjiang Hu;Lincheng Shen
Author_Institution :
College of Mechatronics and Automation, National University of Defense Technology, Changsha, China, 410073
Abstract :
As the development of robot autonomous navigation, visual odometry gains more and more attention in the last few years. It can estimate the egomotion of a robot by analyzing the changes of the on-board camera view. This paper gives an overview of visual odometry and the development of its key technologies, such as feature detection, description and matching/tracking, camera pose estimation. The advantages and disadvantages of different methods are compared and the fundamental principles of them are described in this paper.
Keywords :
"Cameras","Visualization","Feature extraction","Computer vision","Optimization","Estimation"
Conference_Titel :
Information and Automation, 2015 IEEE International Conference on
DOI :
10.1109/ICInfA.2015.7279718