| Summary: | 碩士 === 國立臺灣師範大學 === 機電科技研究所 === 92 === In this paper, we propose a low-complexity zerotree compression technique to multi-channel surveillance systems. Our aims are to decrease the compression time, to raise the compression ratio and to improve the compression quality. Under the developed environment, users can setup different compression ratio, according to different users’ Internet condition, to raise the flexibility of transmitting time and storage space. On the motion detection side, we use the multi-resolution feature of wavelet image on the lower frequency to make the immediate motion detection for lowering the computation complexity and the effect of light and noise. This system, including the input and output of digital image, conducts the Internet communication of Client/ Server through TCP/IP protocol. The system will automatically start Alarm-Trigger when unusual item is detected. Finally, this system is programmed with Borland C++ Builder.
|
|---|
