Somatosensory Input Technology Research Based On Inertia Measurement And SEMG Control

The desire for advanced intellectualization of entertainment, production andlifestyle has risen with rising living standard. On the one hand, Original input methodsuch as keyboard and mouse is not thus enough for interaction demand in terms ofadvanced display technology, internet of things is more and more popular. Varioushousehold appliances, doors and windows become intellectualized, for example,four-rotor air crafts.3D vision and motion sensing games. On the other hand, withthe development of computer technology and transducer technology the basicmanipulation devices, including four-rotor air crafts and so on has appeared more inpeople’s life. So switchover among keyboard, mouse, remotes, and touch screenbecomes more than before, a completely new human-computer interaction deviceapplied for various appliances is being required. Basically, most of the human-computer interaction is conducted by hands, in which the gesture including thecorresponding crook degree and force state of each finger, and the movementinformation of hands including position and posture, are essential in manipulating allkinds of devices. There are two basic gesture import technology that one isdeveloped on the basis of optical instruments, in which the instruments require thewhole moving area of hands without cover, while the other is composed on the forcefeedback device on each finger knuckle, which has been an obstruction of movementof fingers and can not measure the movement information of hands.This paper proposes a new gesture input method, combining inertiameasurement and identification of surface electromyogram to obtain gesture andmovement information. The inertia measurement can obtain corresponding posture,speed and position of the system without the outside reference information, while theidentification of surface electromyogram obtains the state of body from themeasurement of surface electromyogram. The combination of the two technology can get the whole essential information of gesture input, which can also imitate mostof the imput methods of input method and develop various special input method. Sothe potential application prospect of the system completely depends on the precisionthese two technologies.By researching the theory and arithmetic of inertia measurement, the paperdesigns and proposes an inertia measurement system on the basis of micro-electro-mechanical sensors, to obtain the posture, speed and position information and so on,and measures precise posture information from integrated attitude with magneticfield meter. Acquisition circuit of electromyogram signal is proposed based on thecharacteristics of surface electromyogram to realize better effects. Four time domaincharacteristics with neural network selector are applied to establish the identificationsystem of surface electromyogram based on the research of characteristic extractionand selection algorithm of surface electromyogram identification, in order to realizegesture identification in real time. Finally, mouse input method is replaced in actualapplication with the combination of the proposed two systems with proper precision.