Intra prediction method and apparatus for performing adaptive filtering on reference pixel

Abstract

Disclosed is a method for encoding and decoding an image by applying different filtering strength to reference pixels based on a size relationship between a current block and each reference block such that a prediction block for the current block is constructed using reference pixels which are filtered adaptively according to their properties.

1 . A method of decoding an image, with a decoding apparatus, comprising:
 obtaining a flag from a bitstream;   determining, based on the flag, one filter among a plurality of filters pre-defined in the decoding apparatus, the flag indicating the one filter among the plurality of filters;   filtering, based on the determined filter, a reference sample used for prediction of a current block; and   generating a prediction block of the current block based on the filtered reference sample,   wherein the plurality of the filters includes a first filter and a second filter,   wherein a number of taps of the first filter and a number of taps of the second filter are different from each other, and   wherein the first filter is a 5-tap filter.   
     
     
         2 . The method of the  claim 1 ,
 wherein when a value of the flag is 1, the first filter is determined, and   wherein when a value of the flag is 0, the second filter is determined.   
     
     
         3 . The method of the  claim 1 ,
 wherein the filtering of the reference sample is performed using neighboring samples adjacent to the reference sample.   
     
     
         4 . The method of the  claim 3 ,
 wherein the neighboring samples includes at least one of top, bottom, left or right sample of the reference sample.   
     
     
         5 . The method of the  claim 3 ,
 wherein the filtering of the reference sample is performed by applying weights to the reference sample and the neighboring samples.   
     
     
         6 . The method of the  claim 5 ,
 wherein when the first filter is determined, a first weight applied to the reference sample is greater than second weights applied to the neighboring samples,   wherein the first weight applied to the reference sample is greater than three.   
     
     
         7 . The method of the  claim 6 ,
 wherein when the second filter is determined, a first weight applied to the reference sample is two while a second weight applied to at least one of the neighboring samples is one.   
     
     
         8 . A method of encoding an image, with an encoding apparatus, comprising:
 filtering, based on one filter of a plurality of filters pre-defined in the encoding apparatus, a reference sample used for prediction of a current block;   generating a prediction block of the current block based on the filtered reference sample; and   generating a bitstream by encoding the current block based on the prediction block,   wherein a flag is encoded to indicate the one filter among the plurality of the filters,   wherein the plurality of the filters includes a first filter and a second filter,   wherein a number of taps of the first filter and a number of taps of the second filter are different from each other, and   wherein the first filter is a 5-tap filter.   
     
     
         9 . A non-transitory computer readable medium having stored thereon instructions that when executed by a processor perform a decoding method comprising:
 obtaining a flag from a bitstream;   determining, based on the flag, one filter among a plurality of filters pre-defined in the processor, the flag indicating the one filter among the plurality of filters;   filtering, based on the determined filter, a reference sample used for prediction of a current block; and   generating a prediction block of the current block based on the filtered reference sample,   wherein the plurality of the filters includes a first filter and a second filter,   wherein a number of taps of the first filter and a number of taps of the second filter are different from each other, and   wherein the first filter is a 5-tap filter.