High speed audio fingerprint detection to support fair trade of digital contents and real time search.

本研究旨在开发一种超高速音频指纹检测方案，用于从通过互联网交换的音频文件(如mp3)中识别音乐。提出了一种适合硬件实现的基于Haar小波变换的音频指纹合成算法。建议的方案达到了21个。86Gbps的检测速度，几乎比传统的基于FFT的方法快18000倍。

This research addressed to develop an ultra-high speed audio finger print detection scheme to identify music from audio files such as mp3 which are exchanged through internet. We developed a new algorithm to synthesize audio fingerprint based on Haar-wavelet transform which is suitable for hardware implementation. Proposed scheme reached to 21. 86Gbps detection speed and this speed is almost 18 thousand times faster than conventional FFT based method.