Combining Digital Signature with Local Binary Pattern-Least Significant Bit Steganography Techniques for Securing Medical Images

摘要

The image authentication is generally based on two different types of techniques: watermarking and digital signature. In watermarking methods, embedded watermarking is often imperceptible and it contains either a specific ID of producer or codes related to content that are used for authentication. Normally a separate file is stored, digital signature is a non-repudiation and encrypted version of the information extracted from the data. A digital signature can be attached to the data to prove the originality and integrity. The proposed work presents a new approach to steganography of medical images that uses modified Least Significant Bit (LSB) based on the Local Binary Pattern (LBP) pattern. As a first step, cover image has been divided as blocks of 3 x 3 non overlapping masks. Then, the pixel embedding position (clock wise or anti-clock wise) has to be identified using LBP operator. The value of the. LBP operator determines how and where to embed secret image pixel. Later, using LSB method, pixel values will be embedded in the cover image pixel. In order to provide the integrity of the data, the proposed work also presents Reversible Watermarking (RW), a Digital Signature (DS) technique. The proposed algorithm of steganography experimented on few medical images and achieved better efficiency with respect to MSE and PSNR values and same is reported in this paper.