Microphone array method and system, and speech recongnition method and system using the same

PURPOSE: A microphone array method, and a speech recognition method and system using the microphone array method are provided to reduce the number of frequency bins to decrease the quantity of calculations of MUSIC algorithm. CONSTITUTION: A microphone array system includes an input unit(101) using a plurality of microphones for receiving sound signals, a frequency divider for dividing each sound signal input to the input unit into narrow-band signals, and an average spatial covariance matrix estimator(104) for obtaining spatial covariance matrices for sub-arrays and obtaining an average of the spatial covariance matrices. The microphone array system further includes a signal position decision unit for deciding an incident angle of each sound signal through the average covariance matrix, a signal distortion corrector for obtaining a weight to be multiplied by the frequency component of each sound signal based on the incident angle and multiplying the frequency component by the weight, and a signal reproducing unit for reproducing sound signals using the corrected frequency components.