Research Of Audio File Compression Methods Based On Arithmetic Coding

With the development of high technology,the Internet and other emerging fields are also developing rapidly.In this information age,human beings are generating huge amounts of information and data every day.Although the data information contains a lot of valuable information,there is also a lot of redundant information.With the development of the times,people use high technology and the Internet in all aspects to penetrate every part of their lives.Although the network bandwidth is increasing and the hard disk space is expanding,whether it is an open source method of increasing space,or trying to reduce the amount of data.Both of these are the core points to solve the problem.Therefore,how to retain effective information and remove redundant information has become a research hot spot in data processing and the focus of solving huge data problems.Scholars perform a certain re-encoding operation on the data to remove redundant data,and call this process data compression.Although the audio formats commonly used by people,such as MP3,WAV,FLAC and AAC,have been compressed to varying degrees according to the demand,the data still contains a lot of redundant and miscellaneous information.After compressing audio files with the universal software methods such as WINRAR,ZIP,7Z,etc.,we found that the existing algorithms cannot efficiently compress the audio files.Therefore,this article has done some research on the compression of audio files,and based on arithmetic coding,proposed a lossless compression method for audio files.The work of this article is divided into the following two aspects:1.Aiming at the problem that the commonly used compression software is basically invalid for audio files,a coding compression method is proposed.Due to the large number of characters used in the encoding of audio files,it is difficult to use a static encoding model for encoding and compression.At the same time,due to the relatively small correlation between the encoding of audio files,dictionary-based encoding and compression methods not only cannot reduce the amount of data,but also increase the amount of data required for encoding.In this context,this article is based on arithmetic coding,using a Gaussian model to generate the initial value of the symbol frequency for the audio file,the escape code is used to generate the multi-level context model,and the binary index tree is used to store the cumulative frequency of the symbol in the arithmetic coding.Better results.Therefore,the method used in this article can effectively compress audio files without increasing the amount of data after compression.Through a series of experiments,the method proposed in this paper compares with the compression software commonly used in the market,and the compression ratio is better.2.Optimize the encoding and compression according to the format characteristics of audio files.Although the adaptive arithmetic coding method is proved to be effective for the coding and compression of audio files,the coding methods used for audio files of different formats are also completely different.Therefore,we hope to improve the encoding and compression effect of a specific audio file format according to different parts of the file and encoding standards.MP3 audio files have always been the most widely used song format.Therefore,we have conducted a further study on MP3 audio files.For MP3 audio files,we propose an improved RLE algorithm according to its format standard,and through the design and implementation of software modules,we have carried out a compression algorithm for re-encoding.Optimize and improve.Various experimental results prove the effectiveness of the optimization method described in this article.