| Our country has entered the stage of "Population Aging", while the elderly apartment is of low degree of automation and serving ability. With the development of embedded platform, wireless network, computer vision and other technology, it’s feasible of building up a distributed intelligent monitoring system with multiple measurements and wireless transmission.Considering the deficit of PC video monitoring system such as high cost, poor portability, this thesis used TMS320DM6446 from TI company’s DaVinci processor family as its core processor, transmitted information based on ZigBee technology, applied fast image processing algorithm, and realized an indoor distributed intelligent monitoring system. The main work included setting up dual-core software frame, designing and making CC2530 communication interface, improving compressive tracking algorithm, building up a SQLite database and Web server. After test, this system is capable of realizing design goal efficiently.The main contributions and innovations are as below:1. Building up a DaVinci embedded platform. The ARM core operates MontaVista Linux operation system and runs a multi-thread software framework. The DSP core operates modified DSP/BIOS real-time operation system. Communication mechanisms for dual-core embedded platform are based on Codec Engine framework.2. Building up a ZigBee wireless communication network. Communication interface design is based on CC2530 chip. Then the interface board is plated and used for transmitting and receiving data.3. Fulfilling the improvement and implementation of indoor target tracking algorithm. Based on compressive tracking algorithm, the target size changing problem is solved by applying adaptive block strategy, while the under occlusion tracking problem is solved by using block tracking method.4. Setting up the system terminal management. An embedded SQLite database is set up for managing data collected from the apartment. A Web server based on Boa technology is built on PC, while a cellphone app is designed based on Qt technology. Then the system is capable of presenting real-time results on multi-platforms. |