Title :
A Color-Based Gray-Level Compensation Algorithm for Fast Change Detection
Author :
Su, Chang ; Amer, Aishy
Author_Institution :
Electr. & Comput. Eng., Concordia Univ., Montreal, Que.
Abstract :
Gray-level based change detection is effective but suffers when gray-levels of objects are similar to backgrounds. A color-based gray-level compensation algorithm for fast change detection is proposed in this paper. Based on the object scatter estimation in difference frames, the gray-levels of the non-significant pixels in object regions are compensated according to their significance probabilities in color channels. Experimental results show that the proposed method significantly improve the quality of difference frames.
Keywords :
image colour analysis; image resolution; object detection; color-based gray-level compensation algorithm; fast change detection; non-significant pixels; object scatter estimation; Change detection algorithms; Colored noise; Euclidean distance; Image color analysis; Image processing; Image segmentation; Object detection; Scattering; Testing; Video compression; Image processing; image color analysis; image segmentation; object detection;
Conference_Titel :
Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
1-4244-0727-3
Electronic_ISBN :
1520-6149
DOI :
10.1109/ICASSP.2007.366101
