Image decoding device, image decoding method, and program
An image decoding device, includes a filter unit using, as an input, a decoded signal prior to the filtering process and output a filtered decoded signal. The filter unit performs clipping processing on the decoded signal prior to the filtering process such that the absolute value of a differential value between a reference pixel value and a pixel value of the decoded signal prior to the filtering process is less than or equal to a predefined threshold value, and to generate the filtered decoded signal through the linear weighted addition of the value after the clipping processing is performed and the pixel value of the decoded signal prior to the filtering process; and a large-small relationship between the threshold value for a luma signal and the threshold value for a chroma signal are defined such that the large-small relationship is unchanged when an internal bit depth is changed.
1 . An image decoding device, comprising:
a class determination unit configured to use, as an input, a decoded signal prior to a filtering process and determine which class of filter is to be used;
a parameter determination unit configured to determine a filter coefficient for each class; and
a filter unit configured to use, as an input, the decoded signal prior to the filtering process and output a filtered decoded signal, wherein
the filter unit is configured to perform clipping processing on a differential value between a reference pixel value and a pixel value of the decoded signal prior to the filtering process such that an absolute value of the differential value is less than or equal to a predefined threshold value, and to generate the filtered decoded signal through the linear weighted addition of the value after the clipping processing is performed, the filter coefficient and the pixel value of the decoded signal prior to the filtering process,
the threshold value is set for each class and each filter coefficient,
the threshold value for a luma signal is set to be equal to the threshold value for a chroma signal regardless of an internal bit depth, and
the threshold value is defined such that when the internal bit depth is increased by 1, the threshold is doubled.
2 . An image decoding method, comprising:
inputting a decoded signal prior to a filtering process and determining which class of filter is to be used;
determining a filter coefficient for each class; and
using, as an input, the decoded signal prior to the filtering process and outputting a filtered decoded signal, wherein
performing clipping processing on a differential value between a reference pixel value and a pixel value of the decoded signal prior to the filtering process such that an absolute value of the differential value is less than or equal to a predefined threshold value, and the filtered decoded signal is generated through the linear weighted addition of the value after the clipping processing is performed, the filter coefficient and the pixel value of the decoded signal prior to the filtering process,
the threshold value is set for each class and each filter coefficient,
the threshold value for a luma signal is set to be equal to the threshold value for a chroma signal regardless of an internal bit depth, and
the threshold value is defined such that when the internal bit depth is increased by 1, the threshold is doubled.
3 . A program stored on a non-transitory computer readable medium for causing a computer to function as an image decoding device, the image decoding device comprising:
a class determination unit configured to use, as an input, a decoded signal prior to a filtering process and determine which class of filter is to be used;
a parameter determination unit configured to determine a filter coefficient for each class; and
a filter unit configured to use, as an input, the decoded signal prior to the filtering process and output a filtered decoded signal, wherein
the filter unit is configured to perform clipping processing on a differential value between a reference pixel value and a pixel value of the decoded signal prior to the filtering process such that an absolute value of the differential value is less than or equal to a predefined threshold value, and to generate the filtered decoded signal through the linear weighted addition of the value after the clipping processing is performed, the filter coefficient and the pixel value of the decoded signal prior to the filtering process;
the threshold value is set for each class and each filter coefficient,
the threshold value for a luma signal is set to be equal to the threshold value for a chroma signal regardless of an internal bit depth, and
the threshold value is defined such that when the internal bit depth is increased by 1, the threshold is doubled.