Methods and apparatus of prediction refinement using template area

Abstract

A method and apparatus for video coding using prediction refinement are disclosed. According to the method, a template area of the current block is determined. Predicted samples associated with the template area are generated by using an intra prediction process according to the directional mode with reference samples. One or more offsets are derived for a target prediction sample of the current block based on one or more weights, one or more sample differences associated with the template area, or both, wherein said one or more sample differences associated with the template area are calculated between one or more reconstructed samples and one or more predicted samples associated with the template area. The target prediction sample is refined using the one or more offsets to generate a refined prediction sample. The current block is encoded or decoded by using the refined prediction sample.

Images