Robot vision
An HMM-Based Segmentation Method for Traffic Monitoring Movies
IEEE Transactions on Pattern Analysis and Machine Intelligence
Improved Adaptive Gaussian Mixture Model for Background Subtraction
ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 2 - Volume 02
Hi-index | 0.00 |
A spatial-temporal method is proposed for video object detection. The first stage is temporal segmentation which is comprised of GMM on luma and GMM on chroma. An efficient fusion method is proposed to combine the two GMM segmentation results such that the object mask can be improved to some extent. In the second stage, the mask produced in the first stage is statistically analyzed in spatial domain. HMM is employed to refine the segmentation result by estimating the foreground-background state such that the false detection in foreground and background area can be decreased and lead to robust and satisfactory detection results.