Infinity-Norm Rotation for Reversible Data Hiding

In this paper, we propose a novel transform that preserves the dynamic range-infinity-norm rotation. This transform is perfectly reversible and piecewise linear, and keeps the maximum value unchanged. We apply the transform to reversible data hiding, which can be utilized for fragile data hiding and covert communication. After the inverse transform applied to the image with hidden data, no overflow or underflow occurs to the pixel values, and small changes of the coefficients in the transform domain result in small changes of the corresponding pixel values after reconstruction. With progressive symmetrical histogram expansion, we obtain very high embedding capacity for data hiding in the transform domain of infinity-norm rotation. Furthermore, the embedding capacity can be further expanded with blocked coefficients and low-frequency coefficients, which is demonstrated by our experiments.