원문정보
초록
영어
In this paper, we propose a 2D feature tracking method that is stable to long video sequences. To improve the stability of long track-ing, we use trajectory information about 2D features. We predict the expected feature states and compute a rough estimate of the feature lo-cation on the current image frame using the history of previous feature states up to the current frame. A search window is positioned at the esti-mated location and similarity measures are computed within the search window. Once the feature position is determined from the similarity mea-sures, the current feature states are appended to the history bu®er. The outlier rejection stage is also introduced to reduce false matches. Ex-perimental results from real video sequences showed that the proposed method stably tracks point features for long frame sequences.
목차
1 Introduction
2 Improving the Accuracy of Feature Tracking
2.1 Feature extraction
2.2 State modeling for feature tracking
2.3 Rejection of outliers
2.4 Tracking feature points
3 Experimental Results
4 Conclusion
References