A robust image authentication method distinguishing JPEG compression from malicious manipulation

TLDR

Image authentication verifies an image’s originality by detecting malicious manipulations, unlike watermarking, and current methods treat all manipulations equally, whereas practical applications require distinguishing acceptable manipulations such as JPEG compression from malicious ones. The study proposes an image authentication technique that blocks malicious manipulations while permitting JPEG lossy compression. The method uses the invariant relationships between DCT coefficients in corresponding block positions, which survive JPEG quantization, and applies adaptive probabilistic techniques to tolerate distortions from acceptable manipulations such as rounding, filtering, enhancement, and scaling. Experimental and theoretical results show the technique reliably distinguishes malicious manipulations from JPEG compression across all compression ratios and iterations.

Abstract

Image authentication verifies the originality of an image by detecting malicious manipulations. Its goal is different from that of image watermarking, which embeds into the image a signature surviving most manipulations. Most existing methods for image authentication treat all types of manipulation equally (i.e., as unacceptable). However, some practical applications demand techniques that can distinguish acceptable manipulations (e.g., compression) from malicious ones. In this paper, we present an effective technique for image authentication which can prevent malicious manipulations but allow JPEG lossy compression. The authentication signature is based on the invariance of the relationships between discrete cosine transform (DCT) coefficients at the same position in separate blocks of an image. These relationships are preserved when DCT coefficients are quantized in JPEG compression. Our proposed method can distinguish malicious manipulations from JPEG lossy compression regardless of the compression ratio or the number of compression iterations. We describe adaptive methods with probabilistic guarantee to handle distortions introduced by various acceptable manipulations such as integer rounding, image filtering, image enhancement, or scaling-recaling. We also present theoretical and experimental results to demonstrate the effectiveness of the technique.