Method and device for encoding a sub-aperture image of a set of sub-aperture images obtained from a plenoptic image

摘要

A method for encoding a plenoptic image (light field image), comprising a plurality of microlens images the plenoptic image being associated with acquisition parameters of a plenoptic (camera) acquisition system comprises: decomposing the plenoptic image into a plurality of sub-aperture images; and encoding the sub-aperture images using one or more encoding parameters, which encoding parameters are determined in dependence upon the acquisition parameters. The plenoptic image may be classified into one of plural categories by comparing acquisition parameters with thresholds, empirically determined using training data or by statistical modelling of encoding parameter distribution values; category-based encoding using rate-distortion (RD) optimisation or a restricted coding parameter subset may be performed. Sub aperture (sub-images) may be indexed, an encoding order being based on the index. Coding parameters may comprise a search area size for motion estimation (ME), size of blocks of pixels or be related to inter prediction; acquisition (capture, shooting) parameters may comprise a lens aperture or focal length, or focussing plane distance. Also claimed is a similar method and device wherein a subset of possible encoding parameters is determined, for selection of a parameter to use in image coding.