Sensor signal analysis method in dynamic biometric identification systems

The methods used to determine the identity, use a variety of security technology, which main directions are using a variety of sensors to obtain basic data for identification, building architecture, control devices and connect them to your computer. The accuracy of identification of users in such systems is largely determined by the means of obtaining the source data and the formation of a system of signs to identify. When the source data is now priority is being given to the biometric measurements, such as handwriting, voice, etc. The paper considers method of biometrics of handwritten handwriting using multitouch sensor moves to introduce biometric measurements in the form of a multidimensional signal and analysis of each piece of decomposition of the signal as a signal. This approach simplifies the task of markup and analysis of the source of the signal. Biometric image is formed on the basis of dynamic characteristics of multitouch sensor motion signals converted device measuring parameters of refills in a matrix of discrete samples, using Haar transform. Haar transform allows taking into account local characteristics of signals in the form of a transient spikes and provides a dynamic and informative signs.