30-05-2012, 05:29 PM
Sound Steganography - images hidden in spectrogram of music
Sound Steganography - images hidden in spectrogram of music.pdf (Size: 1.36 MB / Downloads: 203)
INTRODUCTION
This work appear in the context of the Mathematical Methods for Measurement
Technology course as the final project were we have chosen a subject of interest and
apply the knowledge in MatLab on its development. For our project we decided to work
with steganography, more specifically sound steganography were we set our objective
as successfully encoding image into sound for later visualization of the image by
reading sound in a frequency spectrum. In order to reach our goal we’ve searched
primarily for information that could point us to an existing work or a starting point for
our project. By this research we realize that the state of the art for sound steganography
by a MatLab approach doesn’t exist or was not published. However, we found some
information regarding the encoding of image into sound done in Perl computer
programming language that give of the base for starting this project.
Steganography
Steganography is defined as the art and science of writing hidden messages in such a
way that the hidden message is unnoticed by someone that is not aware of its content.
However, our project doesn’t fit perfectly in this definition because our goal is to reach
the practical result of encoding of image into sound and then read it again in order to
obtain image trough sound.
MatLab
As development tool we used MatLab that was created by Cleve Moler by 1970s in
University of New Mexico. This powerful tool is capable of numerical computation and
programming in its own language. We applied the use of MatLab in every step of the
project, from reading the picture to the reading of the audio file with a frequency
spectrum from a spectrogram code.
IMPLEMENTATION AND DEVELOPMENT
The project consists in two main parts. The first part is the encoding of the image into
sound and the second part consists in the reading of the sound in a frequency spectrum.
The encoding makes a mapping translation, for each pixel, vertical positioning into
frequency, horizontal position into time-after-click, and brightness into oscillation
amplitude.
Encoding
For the image encoding two other programming functions codes are used. The function
“color” is responsible for the amplitude computation and the function “gen-sin” is
responsible for the wave computation.