Fast pedestrian tracking based on spatial features and colour
Florian H. Seitner, Allan Hanbury · 2006
A tracking with appearance modelling system for pedestrians is described. For pedestrian detection a cascade of boosted classier s and Haar-like rectangular features are used. Statistical modelling in the HSV colour space is used for adaptive background modelling and subtraction, where the use of circular statistics for hue is proposed. A clipping algorithm based on this background model and extensions to the traditional boosted classi- er for fast classication are introduced. By using the background model in combination with the detector, the system extracts a feature vector based on colour statistics and spatial information. Circular and linear statistics are applied on the extracted features to robustly track the ped- estrians and other moving objects. An adaptive appearance model copes with partial or full occlusions and addresses the problem of missing or wrong detections in single frames.