Design And Implementation Of Speaker Recognition System Based On Windows CE

Speaker recognition is the process of automatically recognizing who is speaking on the basis of individual information include in speech signals. It can be widely applied to safety verification, control and so on. Especially, identity Recognition based on telecommunication network. Speaker recognition is a technology that attempts to recognize a speaker through measurements of the specifically individual characteristics arising in speaker's voice. Speaker recognition technology with unique advantages such as convenience, economy and accuracy is paid much attention to. Researching and implementing Speaker Recognition system much more important and significant.The thesis discusses the theory of speaker recognition such as voice pre-processing, feature extraction, recognition algorithms. Linear Prediction Cepstrum Coefficients, Mel-Frequency Cepstral Coefficients and Vector Quantization algorithm are discussed in detail. The thesis improves vector quantization.Linear Prediction Cepstral Coefficients and Vector Quantization are chosen to be implemented in this thesis. The scheme is implemented on Windows CE system. The recognition performance on various system parameters is tested to achieve best scheme. After testing, the system operates correctly, the proposed algorithms have higher accuracy, faster computational speed stronger robustness and the target is reached.