Audio fingerprint extraction method and system thereof

Abstract

The invention provides an audio fingerprint extraction method which comprises the steps of: performing down-sampling processing on original audio data by an audio client for obtaining low-frequency-band audio data, performing quantitative coding processing on the low-frequency-band audio data for obtaining a coding value, and transmitting the coding value to a server; and performing inverse-quantitative coding processing on the received coding value by the server for obtaining reconstructed audio data, and extracting audio fingerprint from the reconstructed audio data. The invention further provides an audio fingerprint extracting system. According to the technical solution provided by the invention, transmission volume of the audio data is reduced and simultaneously accuracy of the audio fingerprint extracted from the audio data is improved.

Description

【Technical field】 [0001] The invention relates to the field of Internet applications, in particular to an audio fingerprint extraction method and system. 【Background technique】 [0002] When the user uses the music player to play local audio or when the user inputs a piece of audio in the music player, the audio fingerprint technology can be used to obtain audio information such as song title, singer, album name, etc. from the audio database of the Internet, and high-level Quality genuine audio resources. In the audio fingerprint technology, it is necessary to compile a feature code for each song to achieve accurate matching and identification of songs, and provide audio identification services for the majority of users. [0003] At present, when using the audio fingerprint technology to obtain audio or audio information from the audio database, the audio client (such as music player software, MP3 player, etc.) needs to upload the compressed audio data to the server through...

AI Technical Summary

Problems solved by technology

The wma encoder is a model-based encoding method. Although it can improve the compression ratio of audio data and reduce the amount of data transmission, it will also lose a lot of audio data, making the accuracy of the extracted audio fingerprints low, which will eventually lead to the identification of the server. The accuracy rate is low; moreover, the encoding process of the wma encoder is very complicated, requiring a large number of calculations and generating a large amount of data, so it cannot be applied in various mobile terminals, which makes the audio client have limitations

Images