An Extended Approach for Color Image Authentication and Recovery Against Composite Attacks

Abstract

This paper proposes an extended approach for color image authentication and recovery utilizing the alpha layer and a secret sharing scheme to counter sequential composite attacks. Building upon our previous work which proved effective against traditional single attacks such as cropping, noise, and superimposition this study addresses the complex challenges posed by successive and combined manipulations. The proposed approach embeds authentication and recovery data within the alpha channel using a spatial correlation phase, ensuring that authentication and repair data are embedded in non-contiguous blocks to guarantee high accuracy in tamper detection and image restoration. Experimental results on diverse color images demonstrate remarkable efficiency in withstanding sequential composite attacks, recording outstanding values of (PSNR=32.85 dB) and (SSIM=0.95) even under the most severe composite sequential attacks. This demonstrates exceptional robustness against tampering while preserving the high visual quality of both the stego and recovered images. Ultimately, the findings confirm that this framework significantly enhances the reliability and security of image transmission and storage compared to conventional methodologies.