Complexity reduction of edge detection for edge based spatial prediction

摘要

The present invention relates to improving the performance of the spatial prediction by reducing the complexity of edge detection for the sub-pel edge based prediction for encoding and decoding of a current image block. In particular, the coding type and/or the coded values of at least one block adjacent to the current block are evaluated. Based thereon it is decided whether an edge detection based on gradient calculation is performed. According to whether the at least one adjacent block includes an edge or not, a type of spatial prediction and/or prediction direction may be selected at the encoder or at the decoder.