An Adaptive Dynamic Background Estimation and Foreground Segmentation for Object Tracking

摘要

Dynamic Background Estimation (DBE) is very effective for moving object tracking when (a part of) background is continually fluctuating. In DBE, newest time-series video frames are kept and the pixel values at the same position in those frames are used to determine the value of background pixels, and one of the major subjects is the parameter adjustment. We propose a new method for DBE, in which we define Mutual Similarity (MS) based on the distance between pixels in a color space, and we choose as the background the pixel that has the highest MS. We also show how the foreground segmentation can be done with the threshold parameter derived from MS calculation. Evaluation results show the effectiveness of our method without any manual parameter adjustment such as using experimental threshold value.