Kieran G. Larkin

Kieran G. Larkin

It is not obvious how to extend Shannon's original information entropy to higher dimensions, and many different approaches have been tried. We replace the English text symbol sequence originally used to illustrate the theory by a discrete, bandlimited signal. Using Shannon's later theory of sampling we derive a new and symmetric version of the second order entropy in 1D. The new theory then naturally extends to 2D and higher dimensions, where by naturally we mean simple, symmetric, isotropic and parsimonious. Simplicity arises from the direct application of Shannon's joint entropy equalities and inequalities to the gradient (del) vector field image embodying the second order relations of the scalar image. Parsimony is guaranteed by halving of the vector data rate using Papoulis' generalized sampling expansion. The new 2D entropy measure, which we dub delentropy, is underpinned by a computable probability density function we call deldensity. The deldensity captures the underlying spatial image structure and pixel co-occurrence. It achieves this because each scalar image pixel value is nonlocally related to the entire gradient vector field. Both deldensity and delentropy are highly tractable and yield many interesting connections and useful inequalities. The new measure explicitly defines a realizable encoding algorithm and a corresponding reconstruction. Initial tests show that delentropy compares favourably with the conventional intensity-based histogram entropy and the compressed data rates of lossless image encoders (GIF, PNG, WEBP, JP2K-LS and JPG-LS) for a selection of images. The symmetric approach may have applications to higher dimensions and problems concerning image complexity measures.

Kieran G. Larkin, Peter A. Fletcher, Stephen J. Hardy

We describe a method for attaching persistent metadata to an image. The method can be interpreted as a template-based blind watermarking scheme, robust to common editing operations, namely: cropping, rotation, scaling, stretching, shearing, compression, printing, scanning, noise, and color removal. Robustness is achieved through the reciprocity of the embedding and detection invariants. The embedded patterns are real onedimensional Mellin monomial patterns distributed over two-dimensions. The embedded patterns are scale invariant and can be directly embedded in an image by simple pixel addition. Detection achieves rotation and general affine invariance by signal projection using implicit Radon transformation. Embedded signals contract to one-dimension in the two-dimensional Fourier polar domain. The real signals are detected by correlation with complex Mellin monomial templates. Using a unique template of 4 chirp patterns we detect the affine signature with exquisite sensitivity and moderate security. The practical implementation achieves efficiencies through fast Fourier transform (FFT) correspondences such as the projection-slice theorem, the FFT correlation relation, and fast resampling via the chirp-z transform. The overall method utilizes orthodox spread spectrum patterns for the payload and performs well in terms of the classic robustness-capacity-visibility performance triangle. Tags are entirely imperceptible with a mean SSIM greater than 0.988 in all cases tested. Watermarked images survive almost all Stirmark attacks. The method is ideal for attaching metadata robustly to both digital and analogue images.