In this paper architecture and all parameters learning algorithm of wavelet-neuro-compressor of large size data for the purpose of detection local features biometric images are proposed. Wavelet-neuro-compressor allows to compress not only images, data in the "object-property" table but and non-stationary nonlinear time series in on-line mode. Proposed approach can be used for solving of the different problems in the intelligent signal processing and the tasks of user authentication with biometrics pattern.
data compression of large size, wavelet-neuro-compressor, biometric pattern, authentication, biometrics
"Vыiavlenye lokalnыkh kharakterystyk byometrycheskykh yzobrazhenyi na osnove vэivlet-neiro-kompressora" [Local characteristics detection of biometric images based on wavelet-neuro-compressor],
Information Processing Systems,