A Codebook-based Video Moving Objects Detecting Method

In this paper, we present a codebook-based background modeling method to detect moving objects from video sequences to adapt to some circumstance changes. The characteristic parameters and numbers of code vector of codebook for each pixel after studied are different, so the background description is more flexible and abundant. The test results show that this algorithm can handle scenes that contain moving background or illumination variation, and it achieves robust detection for compressed videos.