Image Authentication Using Hough Transform Generated Self Signature in DCT Based Frequency Domain (IAHTSSDCT)

Abstract

In this paper a DCT based steganographic technique in frequency domain, termed as IAHTSSDCT has been proposed for authentication of gray scale images. The cover image passes through Hough transformation based on hash function to generate unique signature treated as secret information. As the first step of embedding the cover image transformed into time domain using 2x2 masks in row major order using DCT resulting its corresponding frequency components. Using a secret key and hash function the secret signature/information is embedded into selective AC coefficients. To generate stegoimage those frequency coefficients then passes through inverse DCT. Extraction is also done through same procedure. Experimental results are computed and compared with the existing steganographic techniques like SAWT [1] and YulinWang [3] in terms of Mean Square Error (MSE), Peak Signal to Noise Ratio (PSNR) and Image Fidelity (IF) which show better performances in IAHTSSDCT.