A Scale Invariant Human Motion Detection System using Wavelet Based Feature Extraction

Visual analysis of human motion is one of the most active research topics in computer vision. This strong interest is motivated by a wide range of applications like intelligent video surveillance. There is a need for a motion detection system in which action representation and search scheme shall be able to tolerate enormous variance. In this paper a human motion detection system based on the 2-D wavelet features of history edge images of motion is proposed. This system can effectively encode shape and motion patterns of the image. Furthermore, a booster tree classifier is used for classifying different motions in a scene.