Visual steganalysis of LSB-encoded audio data based on frequency domain characteristics

  • Authors:
  • Yüksel Arslan;Yıldıray Yalman

  • Affiliations:
  • Turgut Ozal University, Ankara, Turkey;Turgut Ozal University, Ankara, Turkey

  • Venue:
  • Proceedings of the 6th International Conference on Security of Information and Networks
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Steganography is an art and science of information hiding. Text, audio and image files are usually used as cover data. Then, a secret message is embedded into these cover data that no one knows about it except sender and receiver. Mostly used methods for audio steganography are based on Least Significant Bit (LSB) method and its derivatives in the literature. Steganalysis is a methodology of detecting messages hidden using steganography. In this paper, a visual steganalysis method has been proposed. Secret text messages can be detected by using frequency domain characteristics of stego audio data which contain text messages. Experimental results show that if stego object contains text messages and it is encoded by LSB or its varieties, some remarkable changes can be observed in frequency domain characteristics and its spectrogram. Thanks to exploring these frequency domain characteristics of stego objects, steganalysis (stego text detection) on audio data can be visually realized.