Dictionary Learning-based, Directional and Optimized Prediction for Lenslet Image Coding
This publication appears in: IEEE Transactions on Circuits and Systems for Video Technology
Authors: R. Zhong, I. Schiopu, B. Cornelis, S. Lu, J. Yuan and A. Munteanu
Publication Date: Apr. 2018
In this paper, a novel approach to encode lenslet images is proposed. The method departs from traditional blockbased coding structures and employs a hexagonal-shaped pixel cluster, called macro-pixel, as elementary coding unit. A novel prediction mode based on dictionary learning is proposed, whereby macro-pixels are represented by a sparse linear combination of atoms from a generic dictionary. Additionally, an optimized linear prediction mode and a directional prediction mode specifically designed for macro-pixels are proposed. Ratedistortion optimization is utilized to select the best intra prediction mode for each macro-pixel. Experimental results on the EPFL light field image dataset show that the proposed coding system outperforms HEVC and the state-of-the-art in lenslet image coding with an average PSNR gain of 3.33 dB and 1.41 dB, respectively, and with rate savings of 67.13% and 34.30%, respectively.