Design Of Smart Home System Based On Somatosensory

With the process of global information, people's lifestyles have also been changed. The survival of residence no longer satisfies with simple needs shelter in the past, but makes more intelligent and more humanized requirements. On the basis of the development of automation technology, computer technology and communication technology, smart home has been rapidly developed in recent years.To achieve household equipment miniaturization and intelligent, this dissertation focuses on designing a smart home system based on somatosensory technology. This paper mainly completes the following work:This paper proposes the entire framework of the smart home system, internal functions, as well as their implementation methods. The system is designed to use MEMS sensors captures gesture, which processes through the somatosensory remote control module and reaches the terminal, through the intelligent home gateway. This process realizes information transfer reliably and effectively. And we elaborate the software and hardware to design of the somatosensory remote control which plays an important role in the smart home system control module, and the intelligent home gateway and the terminal. Finally, we gives the design of the application software. In the design of somatosensory remote control module, a new working mode is proposed, namely switching mode and continuous mode, and realizes a correct detection and switch scheme. In the process of realization of gesture recognition, this paper proposed a new gestures segmentation algorithm, namely double threshold method. This method makes the somatosensory technology better application in the smart home system.The article takes four custom gestures, as an example, tests the performance of gesture recognition system by two ways, which are the user-dependent and the user-independent groups. For user-dependent group, the gesture recognition accuracy is above 95% and the gesture recognition time is within 10 ms. The experimental results and the test data shows that the design can identify and control smart home devices real-time and accurately.The background of the dissertations is the application in market. Because of considering the cost of nodes and power consumption and so on, the design of each module in the system has highly practicability. It can meet the needs of households.