Steganography in webpages using CSS color attributes enhanced by rule-based system

Unlike encryption, steganography's effort is to hide the existence of secret message rather than the meaning of the message only. Web steganography is a newlyborn techniques which are used widely over the web and the internet. However, imperceptibility is one of the most challenging goal in web Steganography techniques. In addition the trade-off between capacity and imperceptibility cannot be denied. Therefore, designing a novel web Steganography approach with a reasonable balance between capacity and imperceptibility, is the main aim in this research. To aim the goal of this research, a novel approach is introduced to steganography in web pages by manipulating the least significant bits of color information in the Cascading Style Sheets (CSS). The overall architecture of the steganography system which includes both the encoder and the decoder at the transmitting and the receiving ends has been described. To achieve the balance relation between capacity and imperceptibility, the system is improved in decision making by rule based expert system. The envisaged system is also secured to ensure both confidentiality and integrity security goals are met. Confidentiality is attained by applying Advanced Encryption Standard (AES) cryptography to the secret message. Integrity is ensured against any modification attempts to the stego-web object through the introduction of Secure Hash Algorithm based checksum (SHA1) in the cookies. The steganography algorithm has been tested subjectively and objectively on twenty web pages and the results demonstrated that this technique is promising which appeared to be of high payload, high imperceptibility and inconspicuous to human observers. The overall system has been tested by six objective and subjective tests such as ACC, PSNR, MSE, chi-square, payload and MOS. The results indicate the steganography technique is highly reliable in terms of imperceptibility and capacity.

Preview Text FK 2016 184 IR.pdf Download (1MB) | Preview

Actions (login required)

View Item