An intelligent video security system using object tracking and shape recognition

Abstract

This paper deals with an intelligent video surveillance system using object tracking and recognition techniques. The proposed system integrates the object extraction, human recognition, face detection, object tracking, and camera control. First, the object in the video frame is extracted using the background subtraction. Then, the object region is examined whether it is human or not. For this recognition, the region-based shape descriptor, angular radial transform, is used to model the human shapes. When the object is decided as human, the face detection is optionally performed to capture the clear face images. Finally, the face or object region is tracked in the video frames, and the pan/tilt/zoom (PTZ) controllable camera also tracks the moving object. The tracking filter updates the histogram information in the object region at every frame so that the moving object is well tracked even though the poses and sizes of object are varied. Since the PTZ parameters can be transformed into camera parameters such as rotation angles and focal length, we estimate the 3-D locations of moving object with multiple PTZ camera. This paper constructs test system with multiple PTZ cameras and their communication protocol. According to the experiments, the proposed system is able to track the moving person automatically not only in the image domain but also in the real 3-D space. The proposed system improves the surveillance efficiency using the usual PTZ cameras. © 2011 Springer-Verlag.