A novel technique for indexing video surveillance data

Abstract

... of video surveillance as a tool for private security, law enforcement and military applications. After appropriate feature extraction has taken place, most video surveillance problems are reduced to the problem of efficiently and robustly matching motion streams. Since all natural motion typically has some variability in the time axis, Dynamic Time Warping (DTW), a technique that aligns the motion streams before calculating their similarity, is typically used. However, DTW can only address the problem of local scaling. As we demonstrate in this work, uniform scaling may be just as important for meaningful automatic analysis of video surveillance data streams. In this work, we demonstrate a novel technique to index of similarity search under uniform scaling. As we will demonstrate, our technique is simple and intuitive, and can achieve a speedup of 2 to 3 orders of magnitude under realistic settings.