Method and apparatus for encoding and decoding image by using large transform unit
A method of encoding an image by transforming a group of prediction units into a frequency domain. A prediction unit may be predicted by using prediction values of the prediction units in the group of prediction units, instead of pixels generated by encoding and then restoring the other prediction units in the group of prediction units. Thus, an encoding compression ratio may be increased as a result of grouping the prediction units.
1. An apparatus for encoding an image, the apparatus comprising:
a predictor that performs inter prediction on a plurality of coding units of the image and determines at least one prediction unit in the coding unit to generate residuals;
a transformer that performs transformation and quantization on the residuals of the coding unit and determines at least one transformation unit in the coding unit, to generate quantized transformation coefficients of the at least one transformation unit; and
an entropy encoder that entropy-encodes the quantized frequency component coefficients,
wherein, when the prediction mode is determined to be an inter prediction mode, not an intra prediction mode, the transformer determines the at least one transformation unit in the coding unit regardless of a size of the at least one prediction unit in the coding unit,
the image is split into a plurality of maximum coding units, according to information about a maximum size of the coding unit,
the maximum coding unit is hierarchically split into one or more coding units of depth including at least one of a current depth and a lower depth, according to split information,
when the split information indicates a split for the current depth, the coding unit of the current depth is split into four rectangular coding units of a lower depth, independently from neighboring coding units, and
when the split information indicates a non-split of the lower depth, one or more prediction units are obtained from the coding unit of the lower depth.
2. The apparatus of claim 1 , wherein the transformer determines a size of the at least one transformation unit based on whether a type of prediction mode is an inter prediction mode or an intra prediction mode.
3. A method of decoding an image, the method comprising:
performing entropy-decoding to obtain quantized transformation coefficients of at least one transformation unit in a coding unit of an image;
performing inverse-quantization and inverse-transformation on the quantized transformation coefficients of the at least one transformation unit to obtain residuals; and
performing inter prediction for at least one prediction unit in the coding unit to generate a predictor and restoring the image by using the residuals and the predictor,
wherein, when a prediction mode is determined to be an inter prediction mode, not an intra prediction mode, the at least one transformation unit in the coding unit is determined regardless of a size of the at least one prediction unit in the coding unit,
the image is split into a plurality of maximum coding units, according to information about a maximum size of the coding unit,
the maximum coding unit is hierarchically split into one or more coding units of depth including least one of a current depth and a lower depth, according to split information,
when the split information indicates a split for the current depth, the coding unit of the current depth is split into four rectangular coding units of a lower depth, independently from neighboring coding units, and
when the split information indicates a non-split of the lower depth, one or more prediction units are obtained from the coding unit of the lower depth.
4. The method of claim 3 , wherein the performing inverse-quantization and inverse-transformation comprises determining a size of the at least one transformation unit based on whether a type of prediction mode is an inter prediction mode or an intra prediction mode.
5. A non-transitory computer readable recording medium having recorded thereon a program for executing the method of claim 3 .
6. An apparatus for decoding an image, the apparatus comprising;
an entropy decoder that performs entropy-decoding to obtain quantized transformation coefficients of at least one transformation unit in a coding unit of the image;
an inverse quantizer that performs inverse-quantization and inverse-transformation on the quantized transformation coefficients of the at least one transformation unit to obtain residuals; and
a restorer that performs inter prediction for at least one prediction unit in the coding unit to generate a predictor and restoring the image by using the residuals and the predictor,
wherein, when a prediction mode is determined to be an inter prediction mode, not an intra prediction mode, the at least one transformation unit in the coding unit is determined regardless of a size of the at least one prediction unit in the coding unit,
an image is split into a plurality of maximum coding units, according to information about a maximum size of the coding unit,
the maximum coding unit is hierarchically split into one or more coding units of depth including at least one of a current depth and a lower depth, according to split information,
when the split information indicates a split for the current depth, the coding unit of the current depth is split into four rectangular coding units of a lower depth, independently from neighboring coding units, and
when the split information indicates a non-split of the lower depth, one or more prediction units are obtained from the coding unit of the lower depth.