Using spatiotemporal blocks to reduce the uncertainty in detecting and tracking moving objects in video

Abstract

We present a novel method for detecting moving objects in videos. The method represents videos using spatiotemporal blocks instead of pixels. Dimensionality reduction is performed to obtain a compact representation of each block's values. The block vectors provide a joint representation of texture and motion patterns. The motion detection and tracking experiments demonstrate that our method although simpler than a state-of-the-art method based on the Stauffer-Grimson Gaussian mixture model has superior performance. It reduces both the instability and the processing time making real-time processing of high resolution videos and efficient analysis of large scale video data feasible.