Sound hiding and extraction using DCT.

number: 
972
English
department: 
Degree: 
Imprint: 
Computer Science
Author: 
Mohammad Nizar Majeed
Supervisor: 
Dr. Loay E. George
Dr. Venus W. Samawi
year: 
2004

Abstract:

Steganography is contemporary way for protecting the data by concealing it into digital media for the purpose of identification, copyright, and send secret massages. This work is concerned with designing and implementing steganographic system that provides six hiding methods with different hiding rates for hiding data in audio signal. The hiding process is mainly done in the transform (frequency) domain by using Discrete Cosine Transformation (DCT) as a transform engine, or hiding may done in the spatial domain by using certain distortion technique, where the hiding in transform domain is described in steganography literature as more robust than the hiding in spatial domain against the attacks. The implemented hiding methods use encryption to encrypt the secret message before hiding it to increase its security. Since the used encryption may or may not require an encryption key, so the proposed hiding methods may considered as two types of steganography: secret key steganography and pure key steganography. Four hiding methods are concerned with hiding an audio (WAV) data in another audio (WAV) data, while the other two hiding methods concerned with hiding any kind of data (like image, text, executable files) in an audio (WAV) data, and all the used audio data is consists of one channel (mono) . The proposed system, also, provides three different fidelity measurement tools (PSNR, SNR, and MAE) to make objective tests; these measurements will assist users in finding the most suitable method for performing the hiding process. It is found that the implemented transform domain hiding methods had a lower SNR and PSNR values and higher MAE than the implemented spatial domain hiding methods, and lower hiding rate.