Method and device for encoding or decoding image
An image decoding method and apparatus according to an embodiment may extract, from a bitstream, a quantization coefficient generated through core transformation, secondary transformation, and quantization; generate an inverse-quantization coefficient by performing inverse quantization on the quantization coefficient; generate a secondary inverse-transformation coefficient by performing secondary inverse-transformation on a low frequency component of the inverse-quantization coefficient, the secondary inverse-transformation corresponding to the secondary transformation; and perform core inverse-transformation on the secondary inverse-transformation coefficient, the core inverse-transformation corresponding to the core transformation.
1. A decoding apparatus comprising:
at least one processor configured to:
obtain first probability information for performing arithmetic decoding for a first bin regarding a first syntax element, based on a first probability model and a second probability model,
obtain the first bin regarding the first syntax element by performing arithmetic decoding based on the obtained first probability information,
obtain a value of the first syntax element corresponding to a first bin string including the first bin,
obtain second probability information for performing arithmetic decoding for a second bin regarding a second syntax element, based on a third probability model and a fourth probability model,
obtain the second bin regarding the second syntax element by performing arithmetic decoding based on the obtained second probability information, and
obtain a value of the second syntax element corresponding to a second bin string including the second bin,
wherein the first probability model is determined based on a first bit-shift value, the second probability model is determined based on a second bit-shift value, and the second bit-shift value is determined to be greater than the first bit-shift value, and
wherein the third probability model is determined based on a third bit-shift value, the fourth probability model is determined based on a fourth bit-shift value, and the fourth bit-shift value is determined to be greater than the third bit-shift value.
2. An encoding method comprising:
obtaining a first bin string by binarizing a value of a first syntax element;
obtaining first probability information for performing arithmetic encoding on a first bin among the first bin string regarding the first syntax element based on a first probability model and a second probability model;
obtaining a first bitstring regarding the first syntax element by performing arithmetic encoding on the first bin based on the obtained first probability information;
obtaining a second bin string by binarizing a value of a second syntax element;
obtaining second probability information for performing arithmetic encoding on a second bin among the second bin string regarding the second syntax element based on a third probability model and a fourth probability model; and
obtaining a second bitstring regarding the second syntax element by performing arithmetic encoding on the second bin based on the obtained second probability information,
wherein the first probability model is determined based on a first bit-shift value, the second probability model is determined based on a second bit-shift value, and the second bit-shift value is determined to be greater than the first bit-shift value, and
wherein the third probability model is determined based on a third bit-shift value, the fourth probability model is determined based on a fourth bit-shift value, and the fourth bit-shift value is determined to be greater than the third bit-shift value.
3. An encoding apparatus comprising:
at least one processor configured to:
obtain a first bin string by binarizing a value of a first syntax element,
obtain first probability information for performing arithmetic encoding on a first bin among the first bin string regarding the first syntax element based on a first probability model and a second probability model,
obtain a first bitstring regarding the first syntax element by performing arithmetic encoding on the first bin based on the obtained first probability information,
obtain a second bin string by binarizing a value of a second syntax element,
obtain second probability information for performing arithmetic encoding on a second bin among the second bin string regarding the second syntax element based on a third probability model and a fourth probability model, and
obtain a second bitstring regarding the second syntax element by performing arithmetic encoding on the second bin based on the obtained second probability information,
wherein the first probability model is determined based on a first bit-shift value, the second probability model is determined based on a second bit-shift value, and the second bit-shift value is determined to be greater than the first bit-shift value, and
wherein the third probability model is determined based on a third bit-shift value, the fourth probability model is determined based on a fourth bit-shift value, and the fourth bit-shift value is determined to be greater than the third bit-shift value.
4. A decoding method comprising:
obtaining first probability information for performing arithmetic decoding for a first bin regarding a first syntax element, based on a first probability model and a second probability model;
obtaining the first bin regarding the first syntax element by performing arithmetic decoding based on the obtained first probability information;
obtaining a value of the first syntax element corresponding to a first bin string including the first bin;
obtaining second probability information for performing arithmetic decoding for a second bin regarding a second syntax element, based on a third probability model and a fourth probability model;
obtaining the second bin regarding the second syntax element by performing arithmetic decoding based on the obtained second probability information; and
obtaining a value of the second syntax element corresponding to a second bin string including the second bin,
wherein the first probability model is determined based on a first bit-shift value, the second probability model is determined based on a second bit-shift value, and the second bit-shift value is determined to be greater than the first bit-shift value, and
wherein the third probability model is determined based on a third bit-shift value, the fourth probability model is determined based on a fourth bit-shift value, and the fourth bit-shift value is determined to be greater than the third bit-shift value.
5. A non-transitory computer readable storage medium storing a bitstream generated by a video encoding method, the bitstream comprising:
a first bitstring regarding a first syntax element; and
a second bitstring regarding a second syntax element,
wherein the video encoding method comprises:
obtaining a first bin string by binarizing a first value of the first syntax element;
obtaining first probability information for performing arithmetic encoding on a first bin among the first bin string of the first syntax element based on a first probability model and a second probability model;
obtaining the first bitstring of the first syntax element by performing arithmetic encoding on the first bin based on the first probability information;
obtaining a second bin string by binarizing a second value of the second syntax element;
obtaining second probability information for performing arithmetic encoding on a second bin among the second bin string of the second syntax element based on a third probability model and a fourth probability model; and
obtaining the second bitstring of the second syntax element by performing arithmetic encoding on the second bin based on the second probability information,
wherein the first probability model has been determined based on a first bit-shift value, the second probability model has been determined based on a second bit-shift value, and the second bit-shift value is greater than the first bit-shift value, and
wherein the third probability model has been determined based on a third bit-shift value, the fourth probability model has been determined based on a fourth bit-shift value, and the fourth bit-shift value is greater than the third bit-shift value.