Biometrics are methods of using the human body (including iris scan, retina scan, finger scan, hand geometry, voice

        verification, and dynamic signature verification) to identify or verify an individual. Applications using biometrics include

        data security, physical access, and customer identification. Biometric systems have quickly become the most popular,

        efficient, secure means of identification in high-security work areas. We selected voice verification due to the fact that all

        other methods require high-dollar input hardware (i.e. fingerprint scanners, retinal scanners, etc…).

                As stated in the Executive Summary, we initially planned to produce a program which took audio input and a

        program which compared and identified voices, but the complexity and time involved in creating the audio-input program

        required that we use an existing program. This program was run on a Microsoft Windows based machine and analyzed

        audio files in WAV format. When a WAV file is analyzed, the set of all Frequency, Phase, Amplitude, and Time points

        were written to a text file. An audio file of 4 seconds has an average of 500,000 total numerical points.

                After we discovered this program we revised our final plan for the voice-recognition program. We realized that we

        could use the Spectrogram program on a Windows 95 computer to analyze a human voice and create a text file which

        would then be analyzed by a program written by our team on the supercomputer and compared with a pre-formatted
 
        database of  "text voices."