Research And Application Of Smart Home System With Kinect Somatosensory Interactive Features

With the rapid development of smart home industry and Smart home-related products have been produced during past two years, the problem that traditional GUI-based smart home systems which is focus on the device functions implemented lead to poor user interactive experience cause more and more attention now. And the next generation of smart home system which is based on NUI somatosensory interactive and focus on user interactive experience enhanced become the trend for the future.In the field of NUI,Kinect is the most potential device.Unlike the previous paper which is focus on the realization of automatic control technology, in this paper, combined with design knowledge, we put the interaction between man and NUI smart home system into two parts—Proactive interaction and Passive interaction.Based on the Kinect sensor, we study the technical aspects of Interactivity in the NUI-based smart home systems as a new perspective.In proactive interaction part, we conducted research and innovation on the proactive identification algorithm which is regarded as the core issues. We applied Kinect-based depth data to gait recognition as the first time, proposed KWU gait recognition algorithm based on the front of the body and dark adaption. It applied depth data to the detection in human movement and gait cycle. During the feature extraction,we extract Fourier descriptors,human height based on depth geometric perspective field and stride length based on depth data.Then serial fused them together to get the gait recognition result.After that, combined with the results of face recognition based on PCA,by the way of put them fused in the decision level,we get the final proactive identification result. We use this result to do a few experiments on the self-built face and gait hybrid database,receiving an achieved good results.Meanwhile we designed a simulation experiments based on smart home scenarios for the algorithm, the results show that the algorithm improves the interactive experience which meet the needs of NUI-based smart home.In passive interaction part,we applied somatosensory interaction into smart home control system in a Innovation way, designed and implemented a smart home care system based on Kinect gesture recognition. At the beginning of this part gave a detailed description on the implement of wave gesture recognition algorithm,then designed a server system based on message registration mechanism supporting multiple users simultaneously online,in the end described the implement of real-time location display system based on Android and Baidu Map.Finally,we conducted experiments on the system and achieved good results.Using Kinect somatosensory interactive features,this paper conducted technology research on the two aspects of interaction of NUI-based smart home system,and have achieved good results finally.