An Adaptive Pixel Selection Method based on Block-Complexity Estimation for Quality Enhanced Steganography

Abstract

This work offers a quality enhanced method of content-adaptive image steganography. The proposed method is divided into three main steps: image segmentation, pixel complexity identification, and data embedding. An input cover image is divided into small local regions and the pixel complexity is identified based on a high pass filter bank and the proposed Least Smoothness Prior (LSP) criterion. Following the criterion, seven complexity levels are defined and a block is assigned from one of the seven levels. The data embedding for the higher complexity levels then takes place using a highly efficient algorithm. Experimental results verify the preservation of visual quality of stego image produced by the proposed method. Two image datasets: SIPI and BOWS2 are used for the experimentation and comparison with prior state-of-art methods. Highest values of the IQ metrics: WPSNR and SSIM show the effectiveness of the proposed algorithm.