Reversible Data Hiding by Coefficient-bias Algorithm
Ching-Yu Yang, Wu-Chih Hu, Chih‐Hung Lin · J. Inf. Hiding Multim. Signal Process. · 2010
A simple lossless data hiding method based on the coefficient-bias algorithm by embedding bits in both spatial domain and frequency domain is proposed. In spatial domain, each pixel in a host image is first subtracted from the block-mean. Then, a stego- image is generated by embedding a large amount of bits (or the primary message) in the mean-removed blocks via the coefficient-bias algorithm. To provide an extra security and robustness, the stego-image is transformed to frequency domain by integer wavelet transform (IWT). A secondary watermark is hidden in the low-high (LH) and high-low (HL) subbands of IWT domain by the proposed algorithm. Simulations show that both the perceptual quality and hiding capacity are not bad. Moreover, the resultant images introduced by the proposed method are tolerant of the attacks such as JPEG2000, JPEG, brightness, and inverting.