Video vehicle detection algorithm through spatio-temporal slices processing

This paper presents new approaches in video vehicle detection. These approaches are developed based upon the pattern analysis of spatio-temporal slices. While traditional approaches to video vehicle detection tend to formulate computational methodologies on the single frame, spatio-temporal slices provide rich information along a larger spatial and temporal scale to improve the precision of vehicle detection. The experiment results demonstrate that the algorithm is real-time, accurate and robust.