Applying homogeneity index modification to high-capacity high-dynamic-range image authentication with distortion tolerance

Distortion-free data hiding algorithms for high-dynamic-range (HDR) images use homogeneity index modification for data embedding. No distortion is encountered between a tone-mapped cover and stego images. However, only approximately one-eighth of pixels have two or more homogeneous representations for data embedding, affecting their performance and applications. This study investigates the feasibility of modifying the homogeneity index for a new application known as HDR image authentication. First, a revision threshold is introduced to ensure that each processing pixel has the maximum number of homogeneous representations possible. In the proposed algorithm, each pixel can perform data embedding. Following that, an authentication code is embedded using a multiple-base notational system. The experimental results indicate that the proposed algorithm significantly increases the embedding capacity and supports HDR image authentication. The data-embedded images with embedded messages are of good quality, and tampered pixels can be effectively detected.

Automated summary

This study proposes a distortion-free data hiding algorithm for high-dynamic-range (HDR) image authentication. By introducing a revision threshold, the proposed algorithm increases the embedding capacity and effectively detects tampered pixels.