A secure LSB ±k steganography based on SFLA

This paper proposes a novel improved LSB ±k steganography based on shuffled frog leaping algorithm(SFLA) In the proposed method, the SFLA is applied to find an optimal substitution matrix for transforming the secret message. Different from the exiting optimization strategies which use the peak signal to noise rate (PSNR) as the fitness to assess the stego image quality, the new algorithm tries to minimize the Markov chain distance to preserve the second-order statistics. The experimental results show that the new algorithm has better security performance than the traditional LSB±k and other improved method under different value of k.