Application of orthogonalization methods of pseudospectra of steps for human identification
A method for processing audio signals of steps of various people in order to effectively separate pseudospectra of signals that can be further subjected to the classification procedure is proposed. The model of a human steps detector was based on the dynamic parameters of the foot position when walking. To record step signals, vibroacoustic sensors were used, sensitive to mechanical and acoustic vibrations in the frequency range from 10 to 1000 Hz. The processing of recorded step signals involved calculating their envelope and then calculating the pseudospectra of the envelope. The procedure for orthogonalization of the pseudospectra of steps for their effective frequency diversity can be used as the basis for the operation of an automated Remote Access Control System. This system can find the application in areas where for one reason or another it is impossible to install another biometric system or is extremely inconvenient, for example, to control access to a room into which the working personnel enter in a special uniform, gloves, and goggles.