Method and apparatus for foreground segmentation of video sequences

摘要

The present invention relates to a method and system for foreground segmentation in which frames of a video sequence are analyzed in the transform domain to determine one or more features. The features are used to model the background. The background can be modeled as a single Gaussian model with a mean and variance of the features. A current frame is segmented by determining if one or more features of the current frame analyzed in the foreground domain satisfy a threshold between the background model. The threshold value can be based on the mean and/or variance of features. During the segmentation, the mean and variance can be updated based on previous corresponding values and current features to adaptively update the background model. In one embodiment, the frames are divided into a plurality of blocks. A transform is used to analyze the blocks in the transform domain. For example, the transform can be a discrete cosine transform (DCT). The features can be a DC feature comprising the DC coefficient and an AC feature comprising a weighted sum of the AC coefficients. The weighted sum is determined with weights which are varied to emphasize different aspects of the present invention. Additional processing steps can be used to remove false positives, handle sudden global illumination changes, handle sudden local illumination changes and remove false negatives.