Secure binary image steganography based on minimizing the distortion on the texture
Download as DOCX, PDF0 likes443 views
The document presents a new binary image steganography scheme that focuses on minimizing distortion in texture to enhance security against steganalysis. It utilizes complement, rotation, and mirroring-invariant local texture patterns to evaluate pixel flipping distortion and develops a practical scheme that partitions images into superpixels and applies syndrome-trellis coding for embedding. Experimental results indicate that the method achieves statistical security while maintaining image quality and capacity.
Secure binary image steganography based on minimizing the distortion on the texture
- 1. SECURE BINARY IMAGE STEGANOGRAPHY BASED ON MINIMIZING THE DISTORTION ON THE TEXTURE ABSTRACT Most state-of-the-art binary image Stefano graphic techniques only consider the flipping distortion according to thehuman visual system, which will be not secure when they are attacked by steg analyzers. In this paper, a binary image Stefano graphic scheme that aims to minimize the embedding distortionon the texture is presented. We extract the complement, rotation,and mirroring-invariant local texture patterns (crmiLTPs)from the binary image first. The weighted sum of crmi LTP changes when flipping one pixel is then employed to measurethe flipping distortion corresponding to that pixel. By testingon both simple binary images and the constructed image dataset, we show that the proposed measurement can well describethe distortions on both visual quality and statistics. Based onthe proposed measurement, a practical stegano graphic scheme is developed. The stegano graphic scheme generates the cover vectorby dividing the scrambled image into superpixels. Thereafter,the syndrome-trellis code is employed to minimize the designedembedding distortion. Experimental results have demonstrated that the proposed stegano graphic scheme can achieve statisticalsecurity without degrading the image quality or the embeddingcapacity.