Image Segmentation in Video Sequences: A Probabilistic Approach
This work addresses the problem of improving vehicle identification and tracking in traffic surveillance systems, representing an incremental advancement over existing techniques.
The paper tackled the problem of accurately segmenting moving objects in video sequences, which traditional background subtraction methods handle poorly for slow-moving objects and shadows, by developing a probabilistic model that classifies each pixel using a mixture-of-Gaussians learned with an incremental EM algorithm, resulting in perfect handling of slow-moving objects and effective shadow elimination.
"Background subtraction" is an old technique for finding moving objects in a video sequence for example, cars driving on a freeway. The idea is that subtracting the current image from a timeaveraged background image will leave only nonstationary objects. It is, however, a crude approximation to the task of classifying each pixel of the current image; it fails with slow-moving objects and does not distinguish shadows from moving objects. The basic idea of this paper is that we can classify each pixel using a model of how that pixel looks when it is part of different classes. We learn a mixture-of-Gaussians classification model for each pixel using an unsupervised technique- an efficient, incremental version of EM. Unlike the standard image-averaging approach, this automatically updates the mixture component for each class according to likelihood of membership; hence slow-moving objects are handled perfectly. Our approach also identifies and eliminates shadows much more effectively than other techniques such as thresholding. Application of this method as part of the Roadwatch traffic surveillance project is expected to result in significant improvements in vehicle identification and tracking.