On the Removal of Steganographic Content from Images

Abstract

Steganography is primarily used for the covert transmission of information even though the purpose can be legitimate or malicious. The primary purpose of this work is to build a firewall which will thwart this transmission. This will be achieved by radiometric and geometric operations. These operations will degrade the quality of cover image. However these can be restored to some extent by a deconvolution operation. The finally deconvolved image is subjected to steganalysis to verify the absence of stego content. Experimental results showed that PSNR and SSIM values are between 35 dB - 45 dB and 0.96, respectively which are above the acceptable range. Our method can suppress the stego content to large extent irrespective of embedding algorithm in spatial and transform domain. We verified by using RS steganalysis, difference image histogram and chi-square attack, that 95 per cent of the stego content embedded in the spatial domain was removed by our showering techniques. We also verified that 100 per cent of the stego content was removed in the transform domain with PSNR 30 dB - 45 dB and SSIM between 0.67-0.99. Percentage of stego removed in both domains was measured by using bit error rate and first order Markov feature.