Correlation-and-bit-aware additive spread spectrum data hiding for Laplacian distributed host image signals

•For the Laplace distributed host signal, obtain the optimal decoder of the ASS scheme.•Analyze the decoding error probability of the ASS scheme.•Propose the new CASS scheme for the Laplace distributed host signal, and analyze its decoding error probability.•Perform simulations to demonstrate the superiority of the proposed CASS scheme over the ASS scheme.•As an extension, discuss the CASS scheme for the generalized Gaussian distribution (GGD) distributed host signal.

Watermarking and data hiding are important tools for copyright protection of digital media contents. To reduce the interference effect of the host signal and improve the decoding performance in the popular additive Spread Spectrum (ASS) embedding scheme, we have proposed a novel correlation-and-bit-aware additive Spread Spectrum (CASS) embedding scheme for the Gaussian distributed host signal. However, in practice, the discrete cosine transform (DCT) coefficients of real images follow the Laplacian distribution. For practical application of CASS to real images in the DCT domain, this paper proposes a CASS data hiding scheme for the Laplacian distributed host image signal. To extract the hidden message bits, we derive an optimal decoder of the ASS scheme for the Laplacian distributed host signal. The bit-error ratio (BER) of ASS and proposed CASS schemes are analyzed in theory. Finally, Monte Carlo simulations and tests on real images are carried out to illustrate the decoding performance of the ASS and proposed CASS schemes. Compared with the traditional ASS scheme, the proposed CASS scheme maintains the simplicity of the decoder, significantly reduces the host effect in data hiding and improves the decoding performance remarkably. As an extension, we discuss the CASS scheme for the generalized Gaussian distribution (GGD) distributed host signal in brief.