Development Of Slight Body Motion Recognition Techniques And Systems Based On Datafusion Of Multiple Sensors

In recent years, the problem of aging is increasingly severe in our country.However, some infrastructures for the elderly are far from perfect, especially theolder group who loss of language ability. We analyze the information that theywant to pass only through the body language. Therefore, we need a newer andnimbler technology to identify the body information pass to us. Development ofslight body motion recognition techniques and systems based on data fusion ofmultisensors play a significant role in it.Based on the analysis of body movements at home and abroad developmentsituation and tendency of research, at the same time, this text advanced amultisenser data acquisition system combined with optical, magnetic, and sportinformation. This system analyzes the body distortion and rotation, makinganalysis more comprehensive. Paper comes from Shanxi provincial scientificand technological projects in the “development of slight body motionrecognition techniques and systems based on data fusion of multisensors”, themain work is as follows: (1) For gesture recognition field was in depth research, micro-movementsof limbs is analyzed to identify future trends, explains the significance of theproject.(2) Introduces the concepts and classifications of multisensor data fusionfor human-computer interaction. Analysis the research field of human motionrecognition, human body model is constructed. Early in the project proposes anoverall design scheme; the project has played a good guide role.(3) It describes the overall design of the system architecture and theInfrared sensor （Kinect sensor） for the internal structure is described. At thesame time, it analyzes Dynamic time warping algorithm analysis and algorithmfor hidden Markov models; output the Kinect sensor’s results.(4) Design for multi-sensor data acquisition module is introduced.Including Triaxial magnetic sensor module, MPU-6050, power supply circuits,AD conversion modules. The build and application of peripheral circuit functionmodel are also been introduced.Detailing all the parameters and standard of thedevice of important part.(5) Do the necessary instructions of algorithm design for multisensor datacollection model. Fuse of the data collection by Euler angles and Quaternion,finally displays the results in the upper computer.