A Reversible Self Watermarking Framework for Integral Images

Integral Imaging (InIm) is a promising three-dimensional (3D) content creation and delivery technique. The simplicity of the method constitutes a straightforward and compact way for acquiring and displaying 3D data. However issues like integrity and authentication of a transmitted InIm are not discussed in the literature. In this paper we propose a jointly optimized InIm fragile self watermarking and compression framework which can be used in order to protect sensitive InIm data from tampering. The main concept is based on a disparity based residual embedding framework that combines increased data embedding capacity with improved compacting properties for the encoded image. The effectiveness of the proposed encoder is evaluated using objective quality metrics over a large InIm dataset.