Arab Educational Information Network

This paper presents a framework for a proposed Sound recognition system to identify sound of human and detect who is the speaker. The presented system is an approach for identification sound through Wavelet and pattern recognition. Primary goals of the proposed system were to identify speaker sound by improve performance results and identify the sound. According to many researches and studies to the acoustic signature of many properties which make them the most discriminating between the dynamic characteristics of the human body in the identification and discrimination of sound, so everyone has a unique sound like one, everyone unique nervous system controls the audio device, everyone has their different sound waves from the rest of humanity. Firstly, the system receives sound from the person and converts it to digital signal. Secondly, it makes some pre-processing on signal to makes it clear by removing noise with the following functions: (I) Input Sound; (II) pre-processing; (III) signal feature extraction using wavelet analysis; (IV) classifying signal. And (V) detect the speaker; this showed in the diagram of the proposed system. (Published abstract)