Apparatus, method, and medium encoding/decoding images through application of psycho-visual characteristics
AssigneeSamsung Electronics Co., Ltd.
InventorsSang-jo Lee, Si-hwa Lee
Life of this patent
Estimating the term…
Abstract
A method, medium, and apparatus encoding and/or decoding an image according to human psycho-visual characteristics. According to a method, medium, and apparatus, error values corresponding to a range in which error values can be recognized because of the human psycho-visual characteristic are selectively encoded.
Claims13
A method of encoding an image based on psycho-visual characteristics, the method comprising:
- calculating an error between a value of an input predetermined pixel, of a plurality of pixels forming an image, and a predicted value of a reference pixel corresponding to the predetermined pixel;
- calculating a difference value between two or more pixels among the plurality of pixels forming the image;
- calculating a threshold value by dividing the calculated difference value by a preset value;
- determining a range of the calculated error to be encoded according to a preset operation determining the range, the preset operation determining the range based on a psycho-visual characteristic based recognition of portions of the calculated error selected to be excluded during encoding, wherein the determining of the range comprises selecting a number of most significant bits of the calculated error which correspond to a value equal to or greater than the calculated threshold;
- and
- encoding the calculated error using only the determined range, by encoding only the selected number of most significant bits for the encoding of the calculated error.
A method of decoding an image encoded based on psycho-visual characteristics, the encoded image including encoded calculated error between a value of an input predetermined pixel, of a plurality of pixels forming the image, and a predicted value of a reference pixel corresponding to the predetermined pixel, with the encoded calculated error including a select range of a calculated error calculated by an encoder that generated the encoded calculated error, and with the range of the encoded calculated error being based on a psycho-visual characteristic based recognition of portions of the calculated error excluded during encoding, the decoding method comprising:
- receiving the encoded calculated error;
- and
- restoring a value corresponding to the unencoded calculated error by synthesizing a remainder of the encoded calculated error and combining the synthesized remainder with the received encoded calculated error,
- wherein the select range of the calculated error is determined by:
- calculating a difference value between two or more pixels among the plurality of pixels forming the image;
- calculating a threshold value by dividing the calculated difference value by a preset value;
- and
- selecting a number of most significant bits of the calculated error which correspond to a value equal to or greater than the calculated threshold.
An apparatus encoding an image based on psycho-visual characteristics, the apparatus comprising:
- an error calculation unit to calculate an error between a value of an input predetermined pixel, of a plurality of pixels forming an image, and a predicted value of a reference pixel corresponding to the predetermined pixel;
- an encoding range determination unit to determine a range of the calculated error to be encoded according to a preset operation determining the range, the preset operation determining the range based on a psycho-visual characteristic based recognition of portions of the calculated error selected to be excluded during encoding, wherein the encoding range determination unit further comprises:
- a calculation unit to calculate a difference value between two or more pixels among the plurality of pixels forming the image;
- a threshold calculation unit to calculate a threshold value by dividing the calculated difference value by a preset value;
- and
- a range determination unit to determine the range by selecting a number of most significant bits of the calculated error which correspond to a value equal to or greater than the calculated threshold;
- and
- an encoding unit to encode the calculated error using only the determined range by encoding only the selected number of most significant bits for the encoding of the calculated error.
Dependencies are inferred from the claim text.