| Autor: |
Xuelong Hu, Wei Fang, Wanpei Chen, Tongyu Jiang, Canjun Qian |
| Předmět: |
|
| Zdroj: |
International Review on Computers & Software; Dec2012, Vol. 7 Issue 7, p3502-3506, 5p, 1 Black and White Photograph, 1 Diagram, 1 Chart |
| Abstrakt: |
Object tracking is a very important research subject, and its achievements have been applied in a wide field. In traditional Mean-shift algorithm only the color features of object was used. When the object has a greater rotation, the object could not be tracked effectively. An improved Mean Shift united with texture was proposed. The sub-image of tracking window was decomposed by the Contourlet transform, and the statistical properties of multi-level Contourlet coefficients were used to constitute a texture feature vector. Then it was weighted and fused with the color histogram feature vector as the final matching feature vector. The experimental results show that this method obtains better accuracy and robustness compared with the traditional Mean-shift based on color probabilistic histogram model on the condition of object rotation. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
