동적 환경에서 강인한 영상특징을 이용한 스테레오 비전 기반의 비주얼 오도메트리
Stereo Vision-based Visual Odometry Using Robust Visual Feature in Dynamic Environment
3
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
Visual odometry is a popular approach to estimating robot motion using a monocular or stereo camera. This paper proposes a novel visual odometry scheme using a stereo camera for robust estimation of a 6 DOF motion in the dynamic environment. The false results of feature matching and the uncertainty of depth information provided by the camera can generate the outliers which deteriorate the estimation. The outliers are removed by analyzing the magnitude histogram of the motion vector of the corresponding features and the RANSAC algorithm. The features extracted from a dynamic object such as a human also makes the motion estimation inaccurate. To eliminate the effect of a dynamic object, several candidates of dynamic objects are generated by clustering the 3D position of features and each candidate is checked based on the standard deviation of features on whether it is a real dynamic object or not. The accuracy and practicality of the proposed scheme are verified by several experiments and comparisons with both IMU and wheel-based odometry. It is shown that the proposed scheme works well when wheel slip occurs or dynamic objects exist.
Abstract
Visual odometry is a popular approach to estimating robot motion using a monocular or stereo camera. This paper proposes a novel visual odometry scheme using a stereo camera for robust estimation of a 6 DOF motion in the dynamic environment. The false results of feature matching and the uncertainty of depth information provided by the camera can generate the outliers which deteriorate the estimation. The outliers are removed by analyzing the magnitude histogram of the motion vector of the corresponding features and the RANSAC algorithm. The features extracted from a dynamic object such as a human also makes the motion estimation inaccurate. To eliminate the effect of a dynamic object, several candidates of dynamic objects are generated by clustering the 3D position of features and each candidate is checked based on the standard deviation of features on whether it is a real dynamic object or not. The accuracy and practicality of the proposed scheme are verified by several experiments and comparisons with both IMU and wheel-based odometry. It is shown that the proposed scheme works well when wheel slip occurs or dynamic objects exist.
참고문헌 (12)
- M. Maimone, "Two Years of Visual Odometry on the Mars Exploration Rovers", Jour. Field Robotics: Special Issue on Space Robotics, 24(3), pp.169~186 (2007)
- Dynamic Stereo Vision, "L. Matthies", Carnegie Mellon Univ., ~ (1989)
- S.Umeyama, "Least-squares estimation of trans-formation parameters between two point patterns", IEEE Trans. Pattern Analysis and Machine Intelligence, pp.376~380 (1991)
- N. Sunderhauf, "Visual odometry using sparse bundle adjustment on an autonomous outdoor vehicle", pp.157~163 (2005), Tagungsband Autonome Mobile Systeme
- K. Konolige, "Large-Scale Visual Odometry for Rough Terrain", International Symposium on Research in Robotics (2007)
- H. Harris, "A Combined Corner and Edge Detector", Proc. of Conf. Alvey Vision, pp.147~151 (1988)
- J.Y.Bouguet, "Pyramidal Implementation of the Lucas Kanade Feature Tracker Description of the algorithm", Intel Corporation, Microprocessor Research Labs (1999)
- "Inc. P. G. R., Triclops Stereo Vision System, Tech. Rep,1997", ~
- I.-K. Kim, "Moving object detection based motion vector and color information", Proc. of Control, Automation, and Systems Symposium (2007)
- R. T. Collins, "Silhouette-based human identification from body shape and gait", Proc. of 5th IEEE Int. Conf. Automatic Face and Gesture Recognition, pp.351~356 (2002)
- I. Pitas, "Digital Image Processing Algorithms" (1993), Prentice Hall
- B.K.P.Horn, "Closed-form solution of absolute orienttation using unit quaternions", Journal of the Optical Society of America A, 4(4), pp.629~642 (1987)