|Table of Contents|

Global Motion Estimation Method with Adaptive Outliers Elimination in Dynamic Scene

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
2011年04期
Page:
442-447
Research Field:
Publishing date:

Info

Title:
Global Motion Estimation Method with Adaptive Outliers Elimination in Dynamic Scene
Author(s):
WANG Xing-mei1YIN Gui-sheng1MEN Zhi-guo2QIU Chen-guang3
1. College of Computer Science and Technology; 2. College of Automation, Harbin Engineering University,Harbin 150001,China; 3. Suzhou Research Centre,East China Research Institute of Photo-electronic,Suzhou 215163,China
Keywords:
dynamic scenes matching feature points global motion estimation outliers
PACS:
TN919. 8
DOI:
-
Abstract:
To exactly obtain global motion estimation in dynamic scene, this paper presents an adaptive global motion estimation method to eliminate outliers. The Best Bin First( BBF) method of the nearest neighbor search algorithm is used to match feature points extracted by the scale invariant feature transform( SIFT) algorithm. In order to improve the accuracy of feature matching,an improved RANdom SAmple Consensus ( RANSAC ) algorithm is proposed that can eliminate outliers adaptively. The iterative number is controlled by the variance of motion magnitude of feature points. Through a camera motion model,accurate results of parameter estimation and background compensation are obtained. The proposed algorithm is tested by the Coastguard standard image sequence and the practical one with dynamic scenes. The experimental results are compared with the previous method,which demonstrates that the proposed algorithm is highly accurate and adaptive and that the speed is faster.

References:

[1] Ross D A,Lim J,Lin R S,et al. Incremental learning for robust visual tracking[J]. International Journal of Computer Vision, 2008, 77( 1-3) : 125-141.
[2] Chum O,Matas J,Obdrzalek S. Enhancing RANSAC by generalized model optimization[A]. Proceedings of the Asian Conference on Computer Vision ( ACCV) [C]. Seoul,South Korea: Asian Federation of Computer Vision Societies, 2004: 812-817.
[3] 陈付幸,王润生. 基于预检验的快速随机抽样一致 性算法[J]. 软件学报, 2005, 16 ( 8) : 1431-1437.
[4] Nister D. Preemptive RANSAC for live structure and motion estimation [ J ]. Machine Vision and Applications, 2005, 16( 5) : 321-329.
[5] Capel D. An effective bail-out test for RANSAC consensus scoring[A]. Proceedings of the British Machine Vision Conference [C]. Britannia,Britain: Oxford, 2005: 1-10.
[6] 田文,王宏远,徐帆,等. RANSAC 算法的自适应Tc, d 预检验[J]. 中国图象图形学报, 2009, 14 ( 5) : 973 -977.
[7] 卓志敏,杨莘元,杨雷. 基于RANSAC+LS 算法的红 外成像全局运动估计[J]. 兵工学报, 2008, 29 ( 3) : 308-312.
[8] Ondrej C,Jirf M. Optimal randomized RANSAC[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30( 8) : 1472-1482.
[9] Tang Chengyuan,Wu Yileh,Hor M,et al. Modified SIFT description for image matching under interfereence [A]. Proceedings of the Seventh International Conference on Machine Learning and Cybernetics[C]. Kunming,China: IEEE Computer Society,2008: 3294 -3300.
[10] Guo Shuxiang,Qiu Chenguang,Ye Xiufen. A kind of global motion estimation algorithm based on feature matching[A]. Proceedings of the 2009 IEEE International Conference on Mechatronics and Automation [C]. Changchun,China: IEEE Computer Society, 2009: 107-111.

Memo

Memo:
-
Last Update: 2012-10-23