Frequency Domain Analysis for Audio Data Forgery Detection

Information security is a classical need of human, from ancient time the sensitive and private data is secured using different kinds of techniques. Now in these days data are becomes digitized and due to use of internet based applications spread world-wide. The available data on internet can easily duplicated or manipulated for re-distribution purpose, this act is known as forgery of digital contents. In this presented work the digital data forgery is investigated and their possible solutions are explored. In observations that are found there are a number of preventive methods are available such as watermarking and copy right acts. But due to diversity of data they become less effective. In this paper the audio data based forgery detection is proposed. For that purpose different audio file formats and their attributes are investigated and wav file format is adopted for analysis. For analysing the audio files Fourier transform for time and frequency domain analysis is performed on files. In addition of that using the cyclic data preservation and similarity computation the similarity between two audio files are computed for detecting the forged audio files. The implementation of the desired audio forgery detection tool is performed using the visual studio environment. Additionally for justification of work the performance of the proposed system is evaluated in terms of time complexity, space complexity, time domain error, frequency domain error and the best overlapped file parts. During the experimentations the performance of the proposed audio file analysis system found optimum and adoptable for forensic usage.