An Embedded DSP Moving Objets Tracking System's Realization

Motion detection and tracking of moving objects is the low-level vision problem. It focuses on the detection and tracking of moving objects in sequences of images. In recent years, moving objects motion analysis is receiving increasing attention from researchers of different fields of study. The interest is motivated by a wide spectrum of applications, such as surveillance.In this thesis, we concentrate on how quickly and efficiently to detect and track moving object in complex environments. And we use active camera to track moving objects continuously. The research includes following aspects.(1) A new real-time algorithm for detecting moving objects is presented. A background subtraction method based on a difference of intensity and edge detection of chromaticity utilized to segment the moving regions. The algorithm is efficient and fit for DSP system.(2) An active vision system for continuous tracking moving objects is presented in this thesis. The moving object can be kept in the center of the camera view by controlling the camera to move with object.(3) According to the hardware architecture of DSP, optimize the 2D-DCT's transfer of JPEG compression on DSP system. The characters of 2D-DCT and Chen's algorithm reduce the cycle counts of the transfer further more. Made a stability foundation for the digital image compression and transition.