Real Time Speech Steganography for Secure Data Transmission

Abstract

Steganography is the art of hiding information into a cover object. The cover object could be any media like an image, audio or human speech. The essence of steganography lies in the ability to hide information in the cover object without degrading its quality and hence giving away hints of tampering. This project involves a spread spectrum representation-based speech steganography using discrete wavelet transform (DWT), which decomposes the cover speech signal into approximated and detail coefficients i.e., low frequency and high frequency components. Our proposed speech steganography provides enhanced imperceptibility since DWT reconstructs the decomposed information without degrading the quality of speech thus staying true to the nature of steganography. Our proposed approach is an extended version of existing Fast fourier transform (FFT) based steganography, where there is a lack of imperceptibility. Compared to existing FFT and any other existing methods it has been observed that there is a lower bit error rate and greater discrepancy. Apart from this we have noticed reduced effects of noise attacks.