Discrete wavelet transform unit and method for adaptively encoding still image based on energy of each block
AssigneeSamsung Electronics Co., Ltd.
InventorsDae-Won Kim, Hyun-Seung Lee, Dae-Kyu Shin
Life of this patent
Estimating the term…
Abstract
Disclosed is a discrete wavelet transform (DWT) unit for encoding and decoding still images based on the energy of each block. The DWT unit includes an energy calculating section for calculating energy of an input image in unit of a block having a predetermined number of pixels, an image decomposing section for performing three-level decomposition of the image of the block by decomposing a band at which a low frequency component is distributed three times, and a coefficient substituting section for substituting 0 (zero) for bands at which a high frequency component is distributed among the three-level decomposed bands.
Claims32
A discrete wavelet transform (DWT) unit for encoding and decoding a still image, comprising:
- an energy calculating section for calculating energy of an input image in a unit of a block having a predetermined number of pixels;
- an image decomposing section for performing three-level decomposition of the image of the block by decomposing a band, at which a low frequency component is distributed three times,
- wherein a range of the calculated energy is divided into three partitioned ranges, such that the three partitioned ranges are respectively assigned into a first energy level, a second energy level and a third energy level in an order of a magnitude of the energy;
- and
- a coefficient substituting section for substituting 0 (zero) for a coefficient of at least one band in which a high frequency component is distributed among the three-level decomposed bands according to the range to which a value of the energy of the corresponding block belongs.
A discrete wavelet transform (DWT) unit for encoding and decoding a still image, comprising:
- an energy calculating section for calculating energy of an input image in a unit of a block having a predetermined number of pixels;
- an image decomposing section for performing adaptive variable-level decomposition of a corresponding block according to a magnitude of energy value calculated in a unit of the block;
- and
- a coefficient substituting section for substituting 0 (zero) for bands, at which a high frequency component is distributed among the adaptive variable-level decomposed bands based on an energy value of the corresponding block and performing image reconstruction.
A method for reconstructing a still image using a discrete wavelet transform (DWT) unit for encoding and decoding the still image, the method comprising:
- a) calculating energy of an input image in a unit of a block having a predetermined number of pixels;
- b) performing three-level decomposition of a band, at which a low frequency component is distributed, relative to the image of the block,
- wherein a range of the calculated energy is divided into three partitioned ranges, such that the three partitioned ranges are respectively assigned into a first energy level, a second energy level and a third energy level in an order of a magnitude of the energy;
- and
- c) substituting 0 (zero) for a coefficient of at least one band in which a high frequency component is distributed among the three-level decomposed bands according to the range to which a value of the energy of the corresponding block belongs.
A method for reconstructing a still image using discrete wavelet transform (DWT) unit for encoding and decoding the still image, the method comprising the steps of:
- calculating energy of an input image in unit of a block of a predetermined number of pixels at an energy calculating section;
- performing adaptive variable-level decomposition of the corresponding block according to a magnitude of energy value calculated in a unit of the block at an image decomposing section;
- and
- substituting 0 (zero) for bands at which a high frequency component is distributed among the adaptive variable-level decomposed bands based on an energy value of the corresponding block and performing image reconstruction at a coefficient substituting section.
Dependencies are inferred from the claim text.