Method and apparatus for inter color component prediction
The present invention provides an image encoding apparatus carrying out inter-color prediction, comprising a residual block acquisition module obtaining a residual block with respect to a first color component and a residual block with respect to a second color component from a difference between an input block and a prediction block; an inter-color component prediction module carrying out inter-color component prediction by generating a residual signal reflecting a difference between a residual block with respect to the first color component and a residual block with respect to the second color component; a transform module generating a transformat coefficient by carrying out transformation with respect to the residual signal; a quantization module generating quantized data by carrying out quantization with respect to the transform coefficient; and an entropy encoding module carrying out entropy encoding by removing statistical redundancy of the quantized data.
1. An image decoding method carrying out inter-color component compensation, comprising:
generating quantized data by receiving a bitstream;
generating a transform coefficient from the quantized data;
generating a residual signal from the transform coefficient;
generating a residual block with respect to a second color component by modifying the residual signal using a residual block of a first color component; and
obtaining a reconstructed block by adding a prediction block with respect to the second color component to the residual block with respect to the second color component,
wherein the generating the residual block with respect to the second color component generates the residual block with respect to the second color component by modifying the residual signal using a value, corresponding to the residual block associated with the first color component, multiplied by a scaling factor, if a predetermined condition is satisfied,
wherein the scaling factor is included in a transform unit in such a way that a sign value of the scaling factor and an absolute value of the scaling factor are separated from each other,
wherein the scaling factor is encoded in consideration of whether the transform unit is larger than a predetermined size, and
wherein the scaling factor is encoded in consideration of whether an intra prediction mode for the second color component is a derived mode in which a first intra prediction mode corresponding to the first color component is identical to a second intra prediction mode corresponding to the second color component when the intra prediction mode is used,
wherein the number of intra prediction modes for the first color component is different from the number of intra prediction modes for the second color component,
wherein the intra prediction modes for the first color component include two non-directional modes and 33 directional modes, and the intra prediction modes for the second color component include the two non-directional modes, a vertical mode, a horizontal mode and the derived mode,
wherein the intra prediction modes for the first color component is labelled by 0, 1, . . . , 34 and the intra prediction modes for the second color component is labelled by 0, 1, . . . , 4,
wherein the generating the residual block with respect to the second color component generates a compensation value using a right shift process for generating the residual block by adding the compensation value to the residual signal,
wherein a bit number corresponding to the right shift process is larger than or equal to a base 2 logarithm of a horizontal size of the predetermined size.
2. The method of claim 1 , wherein the predetermined condition is determined by a flag included in a Picture Parameter Set (PPS).