Hybrid Kernel-Based Template Prediction and Intra Block Copy for Light Field Image Coding

摘要

Light field (LF) technology can capture both spatial and angular information of a 3D scene, and enable new possibilities for digital imaging. However, one problem that occupies an important position to deal with the LF data is the sheer size of data volume. Therefore, in this paper, we propose an effective LF image coding scheme by combining the kernel-based template prediction method and intra block copy method. In the proposed method, intra block copy method is used to predict the unknown blocks where kernel-based template prediction method fails under the Rate-Distortion (RD) based decision mechanism. Experimental results show that the LF data can be efficiently compressed by the proposed method and the execution time in decoder side can also be reduced by about 35% compared to kernel-based template prediction method.