A Large Payload Data Hiding Scheme Using Scalable Secret Reference Matrix

Abstract

When imperceptibility is an issue, data-hiding techniques typically become limited to small payloads. In this paper, a novel data-hiding scheme is proposed that allows embedding large payloads while maintaining high stego-image quality. The scheme utilizes an N×N sub-block for constructing a reference matrix as a secret key that allows the symmetric embedding and extraction of secret data from a grayscale cover image, where N is a positive integer greater than or equal to two. With proper modification, the proposed method can be directly converted to a symmetric cryptosystem. For cases with acceptable stego-image quality (i.e., PSNR > 34), the experimental results showed that the proposed method can embed, on average, four bits per pixel (bpp), a higher hiding capacity than in previous works, and also offers the ability to adjust the capacity by varying N. The number of solutions for a sub-block reaches the factorial of N2, guaranteeing the security of data embedding and extraction. In addition, the proposed method has low computational complexity and can be implemented in a straightforward manner.

Keywords

• flexible embedding capacity
• reference matrix
• steganography