US-20260129200-A1 - IMAGE ENCODING/DECODING METHOD AND DEVICE, AND RECORDING MEDIUM IN WHICH BITSTREAM IS STORED

US20260129200A1US 20260129200 A1US20260129200 A1US 20260129200A1US-20260129200-A1

Abstract

Disclosed herein is an image decoding method. The method of decoding an image comprises obtaining prediction information and residual signal information for a current block from a bitstream and reconstructing, based on the obtained prediction information and the residual signal information, the current block, wherein the prediction information is obtained through entropy decoding, and wherein a context model used for the entropy decoding of the prediction information is determined based on a prediction mode of a neighboring block of the current block.

Inventors

Jin Ho Lee
Jung Won Kang
Ha Hyun LEE
Sung Chang LIM
Hui Yong KIM
Jae Gon Kim
Do Hyeon PARK
Yung Lyul Lee

Assignees

ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE
lNDUSTRY-UNIVERSITY COOPERATION FOUNDATION KOREA AEROSPACE UNIVERSITY
INDUSTRY ACADEMY COOPERATION FOUNDATION OF SEJONG UNIVERSITY

Dates

Publication Date: 20260507
Application Date: 20251223
Priority Date: 20181228

Claims (8)

1 . A method of decoding an image, the method comprising: obtaining prediction information for a current block through entropy decoding; obtaining a prediction block of the current block based on the prediction information; generating a residual block of the current block from a bitstream; and reconstructing the current block based on the prediction block of the current block and the residual block of the current block, wherein the prediction information comprises at least one of IBC (Intra Block Copy) mode information, regular merge mode information, MMVD (Merge mode with Motion Vector Difference) mode information, CIIP (Combined Inter and Intra Prediction) mode information, partition mode information, or subblock merge mode information, and wherein the MMVD mode information is decoded based on the IBC mode information and regular merge mode information, and the partition mode information is decoded based on the regular merge mode information, the CIIP mode information and the subblock merge mode information.
2 . The method of claim 1 , wherein when the regular merge mode information indicates a first value, the MMVD mode information is decoded and when the regular merge mode information indicates a second value, the CIIP mode information or the partition mode information is decoded.
3 . The method of claim 1 , wherein the partition mode information is decoded when all of the IBC mode information, the regular merge mode information, the MMVD mode information, the CIIP mode information and subblock merge mode information are decoded as being inactive.
4 . The method of claim 1 , further comprising: obtaining residual signal information of the current block when a horizontal size or a vertical size of the current block is larger than a pre-determined size which is determined based on a maximum transform block size information that is obtained from the bitstream.
4 . A method of encoding an image, the method comprising: determining prediction information for a current block; generating a prediction block of the current block based on the prediction information for the current block; generating a residual block of the current block based on the prediction block of the current block; and entropy-encoding the prediction information, wherein the prediction information comprises at least one of IBC (Intra Block Copy) mode information, regular merge mode information, MMVD (Merge mode with Motion Vector Difference) mode information, CIIP (Combined Inter and Intra Prediction) mode information, partition mode information, or subblock merge mode information, and wherein the MMVD mode information is encoded based on the IBC mode information and regular merge mode information, and the partition mode information is encoded based on the regular merge mode information, the CIIP mode information and the subblock merge mode information.
5 . The method of claim 4 , wherein the partition mode information is encoded when all of the IBC mode information, the regular merge mode information, the MMVD mode information, the CIIP mode information and subblock merge mode information are encoded as being inactive.
6 . The method of claim 4 , further comprising: generating residual signal information of the current block when a horizontal size or a vertical size of the current block is larger than a maximum transform block size.
7 . A non-transitory computer-readable recording medium for storing encoded image information generated by performing steps of: determining prediction information for a current block; generating a prediction block of the current block based on the prediction information for the current block; generating a residual block of the current block based on the prediction block of the current block; and entropy-encoding the prediction information, wherein the prediction information comprises at least one of IBC (Intra Block Copy) mode information, regular merge mode information, MMVD (Merge mode with Motion Vector Difference) mode information, CIIP (Combined Inter and Intra Prediction) mode information, partition mode information, or subblock merge mode information, and wherein the MMVD mode information is encoded based on the IBC mode information and regular merge mode information, and the partition mode information is encoded based on the regular merge mode information, the CIIP mode information and the subblock merge mode information.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS The present application is a continuation application of U.S. application Ser. No. 18/814,388, filed on Aug. 23, 2024, which is a continuation application of U.S. application Ser. No. 17/419,204, filed on Jun. 28, 2021, now granted U.S. Pat. No. 12,088,809, issued on Sep. 10, 2024, which is the National Stage of International Application No. PCT/KR2019/018575, filed on Dec. 27, 2019, which claims priority to Korean Patent Application No. 10-2018-0172113, filed on Dec. 28, 2018, and Korean Patent Application No. 10-2019-0025711, filed on Mar. 6, 2019, with the Korean Intellectual Property Office, which are incorporated herein by reference in their entirety. TECHNICAL FIELD The present invention relates to a method and apparatus for encoding/decoding an image. Particularly, the present invention relates to a method and apparatus for adaptively performing entropy encoding/decoding using a prediction information of a spatially neighboring block. BACKGROUND ART Recently, the demand for high resolution and quality images such as high definition (HD) or ultra-high definition (UHD) images has increased in various applications. As the resolution and quality of images are improved, the amount of data correspondingly increases. This is one of the causes of increase in transmission cost and storage cost when transmitting image data through existing transmission media such as wired or wireless broadband channels or when storing image data. In order to solve such problems with high resolution and quality image data, a high efficiency image encoding/decoding technique is required. There are various video compression techniques such as an inter prediction technique of predicting the values of pixels within a current picture from the values of pixels within a preceding picture or a subsequent picture, an intra prediction technique of predicting the values of pixels within a region of a current picture from the values of pixels within another region of the current picture, a transform and quantization technique of compressing the energy of a residual signal, and an entropy coding technique of allocating frequently occurring pixel values with shorter codes and less occurring pixel values with longer codes. There are various inter prediction modes for improving prediction performance of an image encoder/decoder. As a result, data for expressing detailed information for performing inter prediction mode and intra prediction mode increases, resulting in a compressed bit amount overhead. The conventional encoder/decoder uses each inter prediction mode indicator according to a sequence in order to express various inter prediction information. However, since a pre-determined sequence does not match the sequence of frequency of the inter prediction mode that actually occurs, binary bits are inefficiently allocated, causing performance loss of encoding/decoding. In addition, in the conventional encoder/decoder, as inter prediction modes that perform similar functions are divided into different indicators, there are problems in that the bit overhead for the indicator is increased, and the occurrence probability is distributed, causing a performance loss of the probability-based entropy coding. DISCLOSURE Technical Problem An object of the present invention is to provide an image encoding/decoding method and apparatus with improved encoding/decoding efficiency. Another object of the present invention is to provide an image encoding/decoding method and apparatus for reducing bit overhead by allocating the smaller bits in order of high occurrence frequency by arranging a series of parsing sequences for expressing inter prediction information in order of occurrence frequency from highest to lowest, in expressing inter prediction mode information. Another object of the present invention is to improve entropy encoding performance by combining similar inter prediction modes and expressing them in one mode to reduce bits and to concentrate the probability of occurrence of prediction modes. Another object of the present invention is to perform entropy encoding/decoding of prediction determination information of current coding block based on prediction information of a spatially neighboring block. Another object of the present invention is to provide a recording medium storing a bitstream generated by an image encoding method/apparatus of the present invention. Technical Solution A method of decoding an image according to an embodiment of the present invention may include obtaining prediction information and residual signal information for a current block from a bitstream and reconstructing, based on the obtained prediction information and the residual signal information, the current block, wherein the prediction information is obtained through entropy decoding, and wherein a context model used for the entropy decoding of the prediction information is determined based on a prediction mode of a neighboring