CN108810556B - Method, device and chip for compressing reference frame - Google Patents
Method, device and chip for compressing reference frame Download PDFInfo
- Publication number
- CN108810556B CN108810556B CN201710295126.XA CN201710295126A CN108810556B CN 108810556 B CN108810556 B CN 108810556B CN 201710295126 A CN201710295126 A CN 201710295126A CN 108810556 B CN108810556 B CN 108810556B
- Authority
- CN
- China
- Prior art keywords
- data block
- reference frame
- compression unit
- data
- prediction residual
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/90—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using coding techniques not provided for in groups H04N19/10-H04N19/85, e.g. fractals
- H04N19/91—Entropy coding, e.g. variable length coding [VLC] or arithmetic coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The disclosure relates to a method, a device and a chip for compressing a reference frame data block. By adopting the method, the reference frame data block is divided into a plurality of compression units, and then each compression unit is compressed, compared with the scheme of compressing the whole reference frame data block, the technical scheme provided by the disclosure further refines the object aimed at by the compression operation, so that the flexibility of the compression operation is improved, the object aimed at by the compression operation is not fixed as the whole reference frame data block any more, but each compression unit in the reference frame data block, and then the prediction method and the entropy coding method can be flexibly selected for each compression unit, so that the flexibility of the compression operation is greatly enhanced.
Description
Technical Field
The present disclosure relates to the field of data compression, and in particular, to a method, an apparatus, and a chip for compressing a reference frame.
Background
The purpose of compressing the reference frame is to reduce the read-write bandwidth and storage space of the reference frame. To facilitate real-time decompression, the compression algorithm of the reference frame is required to satisfy two conditions: 1) the compression unit is easy for random access (random access) and should not be too large; 2) the compression algorithm is simple, fast and real-time.
In the related art, compression algorithms of the reference frame include lossy compression and lossless compression. The lossy compression causes problems of errors and deterioration of image quality, and is therefore not preferable; the lossless compression has low compression rate and high complexity, which is not beneficial to hardware realization, and the storage space is saved too much while the read-write bandwidth is neglected.
Disclosure of Invention
The invention aims to provide a method, a device and a chip for compressing a reference frame data block so as to improve the flexibility of compression operation.
In order to achieve the above object, the present disclosure provides a method of compressing a reference frame, comprising:
one reference frame data block is divided into a plurality of compression units, and each compression unit is compressed.
Optionally, compressing each compression unit comprises:
entropy coding is carried out on the same compression unit by adopting at least two entropy coding modes;
and determining the data stream obtained after the compression unit is compressed according to the bit number of the compression unit subjected to entropy coding by at least two entropy coding modes.
Optionally, after compressing each compression unit, the method further comprises:
determining a rule of a storage unit with a fixed size according to the size of the storage space occupied by one reference frame data block, and dividing the storage space of an external memory into a plurality of storage units;
and storing each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, wherein one total data stream is the sum of data streams obtained after compression units of one reference frame data block are compressed.
Optionally, after compressing each compression unit, the method further comprises:
dividing the storage space of an external memory into a plurality of storage units according to a rule that at least twice of the storage space occupied by one reference frame data block is one storage unit;
and storing each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, wherein one total data stream is the sum of data streams obtained after compression units of one reference frame data block are compressed.
Optionally, entropy encoding the same compression unit by using one of the at least two entropy encoding methods includes:
determining that a maximum value of absolute values of respective prediction residuals in the prediction residual data block of each compression unit is M;
encoding a coding mode of a prediction residual data block of the compression unit with the M being 0, and setting a coded codeword of each prediction residual in the prediction residual data block of the compression unit with the M being 0 to null;
to M>The coding mode of the prediction residual data block of the compression unit of 0 is coded, and M is coded in two cycles>Encoding all prediction residuals of compression unit of 0, wherein an absolute value of less than 2 is represented by an n-bit binary number in a first period of the two periodsn-1And the absolute value of the prediction residual is not less than 2n-1The high n bits of the coded codeword of the prediction residual of (2), in the second of the two periods, representing an absolute value of not less than 2 by using an n-bit binary numbern-1The lower n bits of the encoded codeword of the prediction residual of (1), wherein n is an integer greater than 0.
Optionally, entropy encoding the same compression unit by using one of the at least two entropy encoding methods includes:
respectively taking each compression unit as a unit to be processed, and executing the following steps on each unit to be processed:
dividing the prediction residual data block of the unit to be processed into a plurality of prediction residual data sub-blocks with the same size;
taking each prediction residual error data sub-block as a sub-block to be processed, and executing the following steps on each sub-block to be processed:
determining the maximum value of the absolute value of each prediction residual in the sub-block to be processed;
and coding the bit width of the binary number corresponding to the maximum value, and coding each prediction residual in the sub-block to be processed according to the bit width.
The present disclosure also provides an apparatus for compressing a reference frame, comprising:
a dividing module, configured to divide a reference frame data block into a plurality of compression units;
and the compression module is used for compressing each compression unit.
Optionally, the compression module comprises:
the entropy coding module is used for entropy coding the same compression unit by adopting at least two entropy coding modes;
and the determining module is used for determining the data stream obtained after the compression unit is compressed according to the bit number of the compression unit after entropy coding by at least two entropy coding modes.
Optionally, the apparatus further comprises:
the first memory dividing module is used for determining a rule of a storage unit with a fixed size according to the size of the storage space occupied by the reference frame data block, and dividing the storage space of an external memory into a plurality of storage units;
the first storage module is configured to store each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, where one total data stream is a sum of data streams obtained after compression units of one reference frame data block are compressed.
Optionally, the apparatus further comprises:
the second memory dividing module is used for dividing the storage space of the external memory into a plurality of storage units according to the rule that at least twice of the storage space occupied by one reference frame data block is one storage unit;
and a second storage module, configured to store each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, where one total data stream is a sum of data streams obtained after compression units of one reference frame data block are compressed.
Optionally, the entropy coding module comprises:
a maximum value determining sub-module for determining that a maximum value of absolute values of respective prediction residuals in the prediction residual data block of each compression unit is M;
a first encoding sub-module, configured to encode an encoding mode of the prediction residual data block of the compression unit with M being 0, and set an encoded codeword of each prediction residual in the prediction residual data block of the compression unit with M being 0 to null;
a second coding submodule for pair M>Compression unit of 0And M is coded in two cycles>Encoding all prediction residuals of compression unit of 0, wherein an absolute value of less than 2 is represented by an n-bit binary number in a first period of the two periodsn-1And the absolute value of the prediction residual is not less than 2n-1The high n bits of the coded codeword of the prediction residual of (2), in the second of the two periods, representing an absolute value of not less than 2 by using an n-bit binary numbern-1The lower n bits of the encoded codeword of the prediction residual of (1), wherein n is an integer greater than 0.
Optionally, the entropy coding module comprises:
the first execution submodule is used for respectively taking each compression unit as a unit to be processed, and executing the following steps on each unit to be processed:
dividing the prediction residual data block of the unit to be processed into a plurality of prediction residual data sub-blocks with the same size;
a second execution sub-module, configured to use each prediction residual data sub-block as a sub-block to be processed, and execute the following steps for each sub-block to be processed:
determining the maximum value of the absolute value of each prediction residual in the sub-block to be processed;
and coding the bit width of the binary number corresponding to the maximum value, and coding each prediction residual in the sub-block to be processed according to the bit width.
The present disclosure also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method of compressing reference frame data blocks provided by the present disclosure.
The present disclosure also provides a chip, comprising:
a computer-readable storage medium provided by the present disclosure; and
one or more processors to execute the program in the computer-readable storage medium.
Through the technical scheme, the reference frame data block is divided into the plurality of compression units, and then each compression unit is compressed, compared with the scheme of compressing the whole reference frame data block, the technical scheme provided by the disclosure further refines the object aimed at by the compression operation, so that the flexibility of the compression operation is improved, the object aimed at by the compression operation is not fixed as the whole reference frame data block any more, but each compression unit in the reference frame data block, and then the prediction method and the entropy coding method can be flexibly selected for each compression unit, so that the flexibility of the compression operation is greatly enhanced.
Additional features and advantages of the disclosure will be set forth in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure without limiting the disclosure. In the drawings:
fig. 1 is a flowchart illustrating a method of compressing a reference frame data block according to an example embodiment.
Fig. 2 is a schematic diagram of a reference frame data block of size 8x 4.
Fig. 3 is a schematic diagram of a reference frame data block of size 16x 4.
Fig. 4 is a schematic diagram of dividing a reference frame data block of size 8x 4.
Fig. 5 is a schematic diagram of compression units included in a reference frame data block of a certain size of 8x 4.
Fig. 6A is a schematic diagram of left prediction of a prediction residual.
Fig. 6B is a schematic diagram of up prediction of the prediction residual.
Fig. 6C is a diagram of up-left prediction of prediction residuals.
Fig. 6D is a diagram illustrating up-right prediction of the prediction residual.
Fig. 7 is a schematic diagram of a compressed reference frame data block obtained by using the compression method provided by the embodiment of the present disclosure and stored in an external memory.
Fig. 8 is a diagram illustrating an apparatus for compressing a reference frame data block according to an example embodiment.
FIG. 9 is a schematic diagram of a chip shown in accordance with an example embodiment.
Detailed Description
The following detailed description of specific embodiments of the present disclosure is provided in connection with the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the present disclosure, are given by way of illustration and explanation only, not limitation.
First, a method for compressing a reference frame provided in an embodiment of the present disclosure is explained. Fig. 1 is a flowchart illustrating a method of compressing a reference frame data block according to an example embodiment. As shown in fig. 1, a method for compressing a reference frame provided by an embodiment of the present disclosure includes the following steps:
step S11: dividing a reference frame data block into a plurality of compression units;
step S12: each compression unit is compressed.
The embodiment of the present disclosure provides a method for compressing a reference frame data block, which divides a reference frame data block of any size into a plurality of compression units, and then compresses each compression unit.
That is, in the embodiment of the present disclosure, the compression unit is a basic unit that performs compression and decompression. Each compression operation is performed for a single compression unit, rather than for the entire reference frame data block. Therefore, the flexibility of the compression operation is improved, the object aimed at by the compression operation is not fixed to the whole reference frame data block any more, but each compression unit in the reference frame data block, and then the prediction method and the entropy coding method can be flexibly selected for each compression unit, so that the flexibility of the compression operation is greatly enhanced.
Optionally, dividing one reference frame data block into a plurality of compression units includes:
dividing the reference frame data block into a plurality of lines according to the rule that each line of the brightness data block is one line, and each two lines of the chrominance data block are spliced into one line;
at least one of the plurality of lines is treated as a compression unit.
The size of the reference frame data block may be any, for example: AxB (a is the lateral dimension of the reference frame data block and B is the longitudinal dimension of the reference frame data block), and a and B are both integers greater than 0. Another example is: 8Mx4N, and M and N are integers greater than 0. Reference frame data blocks of size 8Mx4N are advantageous for compression, access, motion estimation, and motion compensation, but the methods of compressing reference frames provided by embodiments of the present disclosure are not limited to reference frame data blocks of size 8Mx 4N. Illustratively, in the case where M is 2 and N is 1, the size of the reference frame data block is 16x 4.
The reference frame data block is composed of a luminance data block and a chrominance data block. Taking a reference frame data block of size 8Mx4N as an example, as shown in fig. 2, in the case where M is 1 and N is 1, the reference frame data block of size 8x4 is composed of a luminance (Y) data block of size 8x4, a chrominance (Cr) data block of size 4x2, and a chrominance (Cb) data block of size 4x 2; as shown in fig. 3, in the case where M is 2 and N is 1, the reference frame data block having a size of 16x4 is composed of a luminance (Y) data block having a size of 16x4, a chrominance (Cr) data block having a size of 8x2, and a chrominance (Cb) data block having a size of 8x 2. In the drawings herein, a luminance (Y) data block is indicated by cross-hatching, a chrominance (Cr) data block is indicated by left-diagonal hatching, and a chrominance (Cb) data block is indicated by right-diagonal hatching.
Dividing a reference frame data block of arbitrary size into a plurality of compression units, comprising the steps of:
first, a reference frame data block of an arbitrary size is divided into a plurality of rows. For a luminance data block in any reference frame data block, each line of the luminance data block can be used as a line; for a chroma data block in any reference frame data block, every two lines of the chroma data block may be spliced to form a line.
Then, at least one line is treated as one compression unit. That is, one line may be used as one compression unit, and a plurality of lines (for example, two lines, three lines, four lines) may be used as one compression unit.
One possible implementation is that for a reference frame data block of size 8Mx4N, each row of the reference frame data block (size 8Mx1) may be treated as a compression unit. For a luminance (Y) data block of size 8Mx4N among reference frame data blocks of size 8Mx4N, each row is a compression unit; for a chroma (Cr) data block with a size of 4Mx2N and a chroma (Cb) data block with a size of 4Mx2N among reference frame data blocks with a size of 8Mx4N, each two lines are spliced to form a compression unit.
For example, taking a reference frame data block with a size of 8Mx4N as an example, as shown in fig. 4, when M is 1 and N is 1, the reference frame data block with a size of 8x4 is divided into 6 rows, where each row of a luminance (Y) data block with a size of 8x4 is the 1 st, 2 nd, 3 rd and 4 th rows of the reference frame data block with a size of 8x4, two rows of a chrominance (Cr) data block with a size of 4x2 are spliced into the 5 th row of the reference frame data block with a size of 8x4, and two rows of a chrominance (Cb) data block with a size of 4x2 are spliced into the 6 th row of the reference frame data block with a size of 8x 4.
As shown in fig. 5, if every two lines of the reference frame data block having the size of 8x4 are taken as one compression unit, the reference frame data block having the size of 8x4 is divided into 3 compression units. Similarly, if each line of the reference frame data block having the size of 8x4 is taken as one compression unit, the reference frame data block having the size of 8x4 is divided into 6 compression units.
The above is a process of dividing the reference frame data block on a line basis. In practical applications, the reference frame data block may also be divided on a column basis. The specific process is similar to the process of dividing the reference frame data block on a line basis, and the description is not repeated here.
After dividing a reference frame data block of an arbitrary size into a plurality of compression units according to the above method, each compression unit is compressed.
And compressing a compression unit, wherein the compression unit comprises a prediction step and an entropy coding step. The embodiments of the present disclosure are not particularly limited as to what prediction mode is used and how the compression unit is predicted. Alternative prediction modes include: left prediction mode, up-left prediction mode, and up-right prediction mode. The method of predicting the compression unit is not limited to the following methods:
predicting each line of brightness data in a compression unit, comprising: for the first row of luminance data in the compression unit, the first luminance data is not predicted, and the other luminance data are predicted by left; for any remaining line of luminance data except the first line of luminance data in the compression unit, the first line of luminance data is predicted by up, and other luminance data can be predicted by any one of a left prediction mode, an up-left prediction mode and an up-right prediction mode;
left prediction is adopted for each row of chrominance data in the compression unit.
For example, for the luminance data of line 1 in fig. 4 or fig. 5, the first luminance data (i.e., 0) is not predicted, and the other luminance data (including 1 to 7) are predicted by left. For the luminance data of the 2 nd to 4 th lines in fig. 4 or fig. 5, the first luminance data in each line of luminance data is predicted by using up prediction, and the remaining luminance data except the first luminance data in each line of luminance data is predicted by using any one of a left prediction mode, an up-left prediction mode and an up-right prediction mode.
Taking the prediction of the 2 nd row of luminance data as an example, the up prediction is adopted for the first luminance data in the 2 nd row of luminance data, and the left prediction is adopted for the remaining luminance data in the 2 nd row of luminance data, as shown in fig. 6A; adopting up prediction for the first luminance data in the 2 nd line of luminance data, and adopting up prediction for the remaining luminance data in the 2 nd line of luminance data, as shown in fig. 6B; adopting up prediction for the first luminance data in the 2 nd line of luminance data, and adopting up-left prediction for the remaining luminance data in the 2 nd line of luminance data, as shown in fig. 6C; for the first luminance data in the 2 nd row luminance data, up prediction is used, and for the remaining luminance data in the 2 nd row luminance data, up-right prediction is used, as shown in fig. 6D.
For the same compression unit, different prediction modes or different prediction methods are adopted, and the obtained prediction residual data blocks are different, so that the same compression unit may correspond to a plurality of prediction residual data blocks. In order to reduce the amount of data transmission, one prediction residual data block may be selected from among prediction residual data blocks corresponding to the same compression unit and transmitted. The criteria for selection were: the sum of the absolute values of the individual prediction residuals in the prediction residual data block is minimal.
Alternatively, to further reduce the amount of data transmission, the number of prediction modes employed may be reduced. Illustratively, for a reference frame data block of size 8Mx4N, only the prediction mode representing the luminance data block is needed for decoding, since the chrominance data block is fixed with left prediction. For the luminance data block, since the first row of the luminance data block fixedly adopts left prediction, only the prediction mode representing the remaining 4N-1 rows of luminance data is needed.
Assuming 4 prediction modes are used, 2 bits are needed to represent 4 selectable prediction modes, and thus the prediction mode for each row of luma data is represented by 2 bits (e.g., row 2 luma data is represented by 00 for left prediction, row 2 luma data is represented by 01 for up prediction, row 2 luma data is represented by 10 for up-left prediction, and row 2 data is represented by 11 for up-right prediction), and thus (4N-1) x2 bits need to be transmitted to represent the prediction modes used for each row in a reference frame data block of size 8Mx 4N.
To reduce the amount of data transmission, the number of prediction modes employed may be reduced. Assuming 2 prediction modes are used, 1 bit is needed to indicate 2 selectable prediction modes, and thus the prediction mode of each row of luminance data is indicated by 1 bit (for example, the 2 nd row of luminance data is indicated by 0 if left prediction is used, and the 2 nd row of luminance data is indicated by 1 if up prediction is used), so that (4N-1) x1 bits need to be transmitted to represent the prediction mode used by each row in the reference frame data block with the size of 8Mx 4N.
No matter what prediction mode is used and how the compression unit is predicted, the prediction residual data block of the compression unit can be obtained. After the prediction residual data block of the compression unit is obtained, entropy coding is performed on the prediction residual data block of the compression unit. Specifically, how to entropy encode the prediction residual data block of the compression unit is not specifically limited in the embodiment of the present disclosure.
Optionally, compressing each compression unit comprises:
entropy coding is carried out on the same compression unit by adopting at least two entropy coding modes;
and determining the data stream obtained after the compression unit is compressed according to the bit number of the compression unit subjected to entropy coding by at least two entropy coding modes.
In the related art, the compression operation is performed on the entire reference frame data block, that is, the entire reference frame data block is subjected to prediction and entropy coding, and then the data stream obtained by compressing the entire reference frame data block is stored in the external memory. In the embodiment of the present disclosure, after dividing the reference frame data block into a plurality of compression units, the compression operation is performed on each compression unit. The compression unit is the target of the compression operation, that is, the prediction and entropy coding can be performed on each compression unit, and then the sum of the data streams obtained by compressing each compression unit of the whole reference frame data block is stored in the external memory as a total data stream.
In order to reduce the size of the storage space occupied by an external memory for one total data stream, the data streams constituting the total data stream may be reduced, that is, the data stream obtained after each compression unit is compressed may be reduced. Therefore, the embodiment of the present disclosure proposes that a compression unit can be encoded by at least two entropy encoding methods, at least two entropy encoding results can be obtained, and then according to the bit numbers of the at least two entropy encoding results, an entropy encoding result with a smaller bit number is selected as a data stream obtained after the compression unit is compressed. Because the data stream obtained after each compression unit is compressed is selected to have a smaller number of bits after comparison, the sum of the data streams obtained after each compression unit of one reference frame data block is also the sum of the data streams with a smaller number of bits, so that the purpose of reducing the size of the storage space occupied by the total data stream obtained after the reference frame data block is compressed in the external memory is achieved.
The disclosed embodiments provide two alternative entropy coding methods, entropy coding each compression unit. The entropy coding method provided by the embodiment of the disclosure comprises the following steps:
determining that a maximum value of absolute values of respective prediction residuals in the prediction residual data block of each compression unit is M;
encoding a coding mode of a prediction residual data block of the compression unit with the M being 0, and setting a coded codeword of each prediction residual in the prediction residual data block of the compression unit with the M being 0 to null;
to M>The coding mode of the prediction residual data block of the compression unit of 0 is coded, and M is coded in two cycles>Encoding all prediction residuals of compression unit of 0, wherein an absolute value of less than 2 is represented by an n-bit binary number in a first period of the two periodsn-1And the absolute value of the prediction residual is not less than 2n-1The high n bits of the coded codeword of the prediction residual of (2), in the second of the two periods, representing an absolute value of not less than 2 by using an n-bit binary numbern-1The lower n bits of the encoded codeword of the prediction residual of (1), wherein n is an integer greater than 0.
With the entropy coding method provided by the embodiment of the present disclosure, it is first necessary to determine the maximum value (denoted as M for convenience of description) of the absolute value of each prediction residual in the prediction residual data block of the compression unit. Considering that the value range of M is wide, for example, M may be equal to 1, 3, 5, 7, etc., M takes different values, and the adopted coding modes are different, and in order to facilitate decoding, the adopted coding mode needs to be encoded first. For example: when M is 0, the coding mode is represented as 00; when M is 1, the encoding mode is represented as 01; when M is 3, the coding mode is denoted as 10; when M is 7, the coding mode is denoted as 11.
In addition to the coding mode, each prediction residual in the prediction residual data block of the compression unit needs to be coded. The code word for each prediction residual differs depending on whether M is equal to 0 or not.
If M is 0, only the coding mode needs to be coded, and each prediction residual does not need to be coded, and accordingly, only the coded codeword of the coding mode needs to be transmitted.
If M is>0, the coding mode needs to be coded and each prediction residual is coded with an n-bit binary number. In particular, for absolute values less than 2n-1The prediction residual of (2) is encoded by using one cycle and is encoded into n bits; for absolute value not less than 2n-1The prediction residual of (2) is encoded into 2n bits by using two cycles, wherein the high n bits of the 2n bits are obtained in the first cycle of the two cycles, and the low n bits of the 2n bits are obtained in the second cycle of the two cycles.
For example, when n is 1, M is 1, and thus the prediction residual may be: 0. +1, -1, where the absolute value of the prediction residual 0 is less than 21-1Thus, the prediction residual 0 is encoded with one cycle, which is: 1; absolute values of +1 and-1 of prediction residuals are not less than 21-1Encoding the data into 2 bits by using two periods, and obtaining the high order bits in the 2 bits in the first period of the two periods: 0, in the second of the two cycles, the lower of the 2 bits is obtained: 0 and 1. The code word for which prediction residual 0 is obtained is 1, the code word for prediction residual +1 is 00, and the code word for prediction residual-1 is 01.
For example, when n is 2, M is 2 or 3, if M is 2, the prediction residual may be: 0. +1, -2, + 2; if M is 3, the prediction residual may be: 0. +1, -2, +2, -3, + 3. The prediction residual may also be-3 and +3 in the case of M-3, compared to the case of M-2. Since the case where M is 2 is included in the case where M is 3, M is 3 as an example.
Absolute value of prediction residual 0, +1, -1 is less than 22-1Thus, the prediction residual 0 is encoded with one cycle, which is: 01. 10, 11; absolute value of prediction residuals-2, +2, -3, +3 is not less than 22-1Encoding it with two cycles, encoding it into 4 bits, in the first of the two cycles, obtaining the high 2 bits of the 4 bits: 00, in the second of the two cycles, the lower 2 bits of the 4 bits are obtained: 10. 11, 00 and 01. Thus, the coding code of prediction residual 0 is obtainedWord is 01, the code word for prediction residual +1 is 10, the code word for prediction residual-1 is 11, the code word for prediction residual + 2 is 0010, the code word for prediction residual-2 is 0011, the code word for prediction residual +3 is 0000, and the code word for prediction residual-3 is 0001. In the same manner, the encoding codeword table shown in table 1 can be obtained. In Table 1, "-" indicates null.
TABLE 1 encoding code word table
By adopting the entropy coding method provided by the embodiment of the disclosure, the coding of each prediction residual can be completed in two periods, which is convenient and fast.
Another entropy coding method provided by the embodiment of the present disclosure includes the following steps:
respectively taking each compression unit as a unit to be processed, and executing the following steps on each unit to be processed:
dividing the prediction residual data block of the unit to be processed into a plurality of prediction residual data sub-blocks with the same size;
taking each prediction residual error data sub-block as a sub-block to be processed, and executing the following steps on each sub-block to be processed:
determining the maximum value of the absolute value of each prediction residual in the sub-block to be processed;
and coding the bit width of the binary number corresponding to the maximum value, and coding each prediction residual in the sub-block to be processed according to the bit width. In order to reduce the amount of data transmission, the embodiments of the present disclosure provide that the prediction residual data block of each compression unit is divided into a plurality of prediction residual data sub-blocks having the same size. Each sub-block of prediction residual data is then encoded. If the absolute value of the prediction residual included in one prediction residual data sub-block is smaller, the sub-block is represented by binary numbers with fewer bits; if a sub-block of prediction residual data comprises a prediction residual having a larger absolute value, it is characterized by a more bit binary number. Compared with the method for representing each prediction residual in the prediction residual data block of the compression unit by using the binary number with the same bit, by using the method for coding after grouping provided by the embodiment of the disclosure, a certain group or a certain group can be represented by using less binary numbers, thereby reducing the data transmission amount. Specifically, how to divide the compression unit, the compression device and the decompression device may negotiate in advance, which is not limited in the embodiment of the present disclosure.
Illustratively, for a prediction residual data block of a compression unit of size 8Mx1, one possible grouping method is: each 8 prediction residuals are grouped into M groups, i.e., M sub-blocks of prediction residual data, and the prediction residual data blocks of the compression unit with the size of 8Mx1 are further divided into M groups.
After obtaining a plurality of prediction residual data sub-blocks according to the method, coding each prediction residual in each prediction residual data sub-block according to the maximum value of the absolute value of each prediction residual in each prediction residual data sub-block.
Firstly, determining the maximum value of the absolute value of each prediction residual error in the sub-block to be processed, and further determining the bit width of the binary number corresponding to the maximum value. And then, coding each prediction residual in the sub-block to be processed according to the bit width.
Specifically, assuming that the maximum value of the absolute values of the prediction residuals in the sub block to be processed is a, the bit width of the binary number corresponding to the maximum value may be represented as a
Wherein the content of the first and second substances,
indicating rounding down, then can utilize
Representing each prediction residual in the subblocks to be processed by using bit binary numbers; alternatively, it is possible to utilize
Bit binary number representationEach prediction residual in the sub-block to be processed, wherein,
indicating rounding up.
Illustratively, the maximum value of the absolute value of each prediction residual in a sub-block of prediction residual data is 7, and thus the bit width of the corresponding binary number is 3, so each prediction residual in the sub-block of prediction residual data can be encoded using a 4-bit binary number, wherein,
in a specific implementation process, the method for entropy coding the prediction residual data block of the compression unit is not limited to the two entropy coding methods provided in the embodiment of the present disclosure, and other entropy coding methods may also be used to entropy code the prediction residual data block of the compression unit.
Alternatively, in order to reduce the size of the storage space occupied by the external memory by the sum of the data streams (i.e., the total data stream) obtained after the compression units of the entire reference frame data block are compressed, after entropy encoding the prediction residual data block of the compression unit by using multiple entropy encoding methods including the two entropy encoding methods provided in the embodiments of the present disclosure, the data stream obtained after the compression unit is compressed may be selected from the encoding results corresponding to the multiple entropy encoding methods, where the encoding result has the smallest number of bits. Because the data stream obtained after each compression unit is compressed is selected to have a smaller number of bits after comparison, the sum of the data streams obtained after each compression unit of one reference frame data block is also the sum of the data streams with a smaller number of bits, so that the purpose of reducing the size of the storage space occupied by the total data stream obtained after the reference frame data block is compressed in the external memory is achieved.
After the above prediction step and entropy coding step are performed on a compression unit, the compression operation on the compression unit is completed. Then, the data stream compressed by the compression unit is stored in the external memory.
Thus, after compressing each compression unit, the method further comprises:
determining a rule of a storage unit with a fixed size according to the size of the storage space occupied by one reference frame data block, and dividing the storage space of an external memory into a plurality of storage units;
and storing each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, wherein one total data stream is a sum of data streams obtained after compression units of one reference frame data block are compressed.
The embodiment of the disclosure provides that a fixed-size storage unit is determined according to the size of the storage space occupied by the reference frame data block, the storage space of the external memory is divided, and the storage space of the external memory is divided into a plurality of fixed-size storage units.
Since the storage unit is of a fixed size, before each compression unit included in the reference frame data block is compressed, the storage location of the total data stream obtained by compressing the reference frame data block in the external memory can be predetermined. The total data stream obtained by compressing the reference frame data block is the sum of the data streams obtained by compressing each compression unit of the reference frame data block.
Optionally, after compressing each compression unit, the method further comprises:
dividing the storage space of an external memory into a plurality of storage units according to a rule that at least twice of the storage space occupied by one reference frame data block is one storage unit;
and storing each total data stream into the external memory according to a rule that a total data stream obtained by compressing one reference frame data block occupies one storage unit, wherein one total data stream is a sum of data streams obtained by compressing the compression units of one reference frame data block.
In order to make the 8-bit data stream and the 10-bit data stream compatible, the disclosed embodiment proposes that the external memory is divided into a plurality of memory cells according to at least twice of the memory space occupied by the reference frame data block, for example: the size of one memory cell is twice the space occupied by the reference frame data block. And then, storing a total data stream obtained by compressing a reference frame data block into a storage unit.
Illustratively, a reference frame data block with a size of 8Mx4N is typically an 8-bit data stream, occupying a maximum of 8Mx4Nx3/2 storage space, and if a reference frame data block with a size of 8Mx4N is a 10-bit data stream, occupying a maximum of 8Mx4Nx3/2x10/8 storage space. And 8Mx4Nx3/2 and 8Mx4Nx3/2x10/8 are both smaller than 8Mx4Nx2, so that the external memory can be divided into a plurality of memory cells of size 8Mx4N x2 according to the rule that the memory space of size 8Mx4Nx2 is one memory cell, so that after the compression operation on the reference frame data block of size 8Mx4N is completed, the compressed reference frame data block is stored in the memory cell.
Fig. 7 is a schematic diagram of a compressed reference frame data block obtained by using the compression method provided by the embodiment of the present disclosure and stored in an external memory. In fig. 7, a memory cell with a size of 8Mx4Nx2 is between every two adjacent cursors. Fig. 7 illustrates, for example, a reference frame data block having a size of 8Mx4N, the reference frame data block after compression being represented by dot padding, and the reference frame data block not compressed being represented by black padding. The size of the compressed reference frame data block may be: 16MN, 32MN or 48 MN. Therefore, the four compression modes can be represented by 2-bit binary numbers, and the compression mode comparison table shown in table 2 is obtained.
TABLE 2 compression mode LUT
| Compression mode | Size of compressed reference frame data block |
| 00 | <＝16MN |
| 01 | <＝32MN |
| 10 | <＝48MN |
| 11 | >48MN |
Based on the same inventive concept, the embodiment of the present disclosure further provides a device for compressing the reference frame data block. Fig. 8 is a diagram illustrating an apparatus for compressing a reference frame data block according to an example embodiment. As shown in fig. 8, the apparatus 100 includes:
a dividing module 101, configured to divide a reference frame data block into a plurality of compression units;
a compressing module 102, configured to compress each compressing unit.
Optionally, the compression module comprises:
the entropy coding module is used for entropy coding the same compression unit by adopting at least two entropy coding modes;
and the determining module is used for determining the data stream obtained after the compression unit is compressed according to the bit number of the compression unit after entropy coding by at least two entropy coding modes.
Optionally, the apparatus further comprises:
the first memory dividing module is used for determining a rule of a storage unit with a fixed size according to the size of the storage space occupied by the reference frame data block, and dividing the storage space of an external memory into a plurality of storage units;
the first storage module is configured to store each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, where one total data stream is a sum of data streams obtained after a plurality of compression units of one reference frame data block are compressed.
Optionally, the apparatus further comprises:
the second memory dividing module is used for dividing the storage space of the external memory into a plurality of storage units according to the rule that at least twice of the storage space occupied by one reference frame data block is one storage unit;
a second storage module, configured to store each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, where one total data stream is a sum of data streams obtained after multiple compression units of one reference frame data block are compressed.
Optionally, the entropy coding module comprises:
a maximum value determining sub-module for determining that a maximum value of absolute values of respective prediction residuals in the prediction residual data block of each compression unit is M;
a first encoding sub-module, configured to encode an encoding mode of the prediction residual data block of the compression unit with M being 0, and set an encoded codeword of each prediction residual in the prediction residual data block of the compression unit with M being 0 to null;
a second coding submodule for pair M>The coding mode of the prediction residual data block of the compression unit of 0 is coded, and M is coded in two cycles>Encoding all prediction residuals of compression unit of 0, wherein an absolute value of less than 2 is represented by an n-bit binary number in a first period of the two periodsn-1And the absolute value of the prediction residual is not less than 2n-1The high n bits of the coded codeword of the prediction residual of (2), in the second of the two periods, representing an absolute value of not less than 2 by using an n-bit binary numbern-1The lower n bits of the encoded codeword of the prediction residual of (1), wherein n is an integer greater than 0.
Optionally, the entropy coding module comprises:
the first execution submodule is used for respectively taking each compression unit as a unit to be processed, and executing the following steps on each unit to be processed:
dividing the prediction residual data block of the unit to be processed into a plurality of prediction residual data sub-blocks with the same size;
a second execution sub-module, configured to use each prediction residual data sub-block as a sub-block to be processed, and execute the following steps for each sub-block to be processed:
determining the maximum value of the absolute value of each prediction residual in the sub-block to be processed;
and coding the bit width of the binary number corresponding to the maximum value, and coding each prediction residual in the sub-block to be processed according to the bit width.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
Based on the same inventive concept, the embodiment of the disclosure also provides a chip. Fig. 9 is a block diagram illustrating a chip 700 according to an example embodiment. As shown in fig. 9, the chip 700 may include: a processor 701, a memory 702, and a bus 705. The bus 705 is used to realize data transmission between the chip 700 and an external memory.
The external memory is used for storing the total data stream obtained after the reference frame data block is compressed.
The processor 701 is configured to complete all or part of the steps in the method for compressing the reference frame data block. The memory 702 is used to store various types of data to support the operation of the chip 700, and these data may include, for example, instructions for any application or method operating on the chip 700. The Memory 702 may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk. In an exemplary embodiment, the chip 700 may be a separate chip with compression function, or may be integrated in a codec as a component with compression function in the codec.
In another exemplary embodiment, a computer readable storage medium, such as the memory 702, is also provided that includes program instructions executable by the processor 701 of the chip 700 to perform the above-described method of compressing a reference frame data block.
The preferred embodiments of the present disclosure are described in detail with reference to the accompanying drawings, however, the present disclosure is not limited to the specific details of the above embodiments, and various simple modifications may be made to the technical solution of the present disclosure within the technical idea of the present disclosure, and these simple modifications all belong to the protection scope of the present disclosure.
It should be noted that, in the foregoing embodiments, various features described in the above embodiments may be combined in any suitable manner, and in order to avoid unnecessary repetition, various combinations that are possible in the present disclosure are not described again.
In addition, any combination of various embodiments of the present disclosure may be made, and the same should be considered as the disclosure of the present disclosure, as long as it does not depart from the spirit of the present disclosure.
Claims (9)
1. A method of compressing a reference frame, comprising:
dividing a reference frame data block into a plurality of compression units;
compressing each compression unit;
the reference frame data block comprises a luminance data block and a chrominance data block, and is divided in the following way: taking each row of the brightness data block as a target row, and splicing every two rows of the chrominance data block as a target row to obtain a plurality of target rows; at least one target line in the plurality of target lines is taken as a compression unit;
wherein the compressing each compression unit comprises a prediction step:
for each of the compression units, performing the following steps to implement the predicting step:
adopting a 1eft prediction mode for each row of chrominance data in the compression unit; for the first line of brightness data in the compression unit, the first brightness data is not predicted, and other brightness data adopt a 1eft prediction mode; respectively predicting any remaining line of luminance data except the first line of luminance data in the compression unit by adopting an up prediction mode for the first luminance data in the any line of luminance data and respectively predicting other luminance data except the first luminance data in the any line of luminance data as a whole by adopting a 1eft prediction mode, an up-1eft prediction mode and an up-right prediction mode to obtain different prediction residual data blocks of the compression unit;
the method further comprises the following steps:
and selecting one prediction residual data block from the prediction residual data blocks different from the compression unit for transmission, and coding and transmitting each prediction mode corresponding to the brightness data of any remaining row except the first row corresponding to the selected prediction residual data block.
2. The method of claim 1, wherein after compressing each compression unit, the method further comprises:
determining a rule of a storage unit with a fixed size according to the size of the storage space occupied by one reference frame data block, and dividing the storage space of an external memory into a plurality of storage units;
and storing each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, wherein one total data stream is the sum of data streams obtained after compression units of one reference frame data block are compressed.
3. The method of claim 1, wherein after compressing each compression unit, the method further comprises:
dividing the storage space of an external memory into a plurality of storage units according to a rule that at least twice of the storage space occupied by one reference frame data block is one storage unit;
and storing each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, wherein one total data stream is the sum of data streams obtained after compression units of one reference frame data block are compressed.
4. The method of claim 1, wherein compressing each compression unit further comprises an entropy encoding step, the entropy encoding step comprising:
determining that a maximum value of absolute values of respective prediction residuals in the prediction residual data block of each compression unit is M;
encoding a coding mode of a prediction residual data block of the compression unit with the M being 0, and setting a coded codeword of each prediction residual in the prediction residual data block of the compression unit with the M being 0 to null;
to M>The coding mode of the prediction residual data block of the compression unit of 0 is coded, and M is coded in two cycles>Encoding all prediction residuals of the compression unit of 0, wherein an absolute value smaller than that in a first period of the two periods is represented by an n-bit binary number
And the absolute value of
In the second period of the two periods, the absolute value is represented by n-bit binary numbers not less than
The lower n bits of the encoded codeword of the prediction residual of (1), wherein n is an integer greater than 0.
5. The method of claim 1, wherein compressing each compression unit further comprises an entropy encoding step, the entropy encoding step comprising:
respectively taking each compression unit as a unit to be processed, and executing the following steps on each unit to be processed:
dividing the prediction residual data block of the unit to be processed into a plurality of prediction residual data sub-blocks with the same size;
taking each prediction residual error data sub-block as a sub-block to be processed, and executing the following steps on each sub-block to be processed:
determining the maximum value of the absolute value of each prediction residual in the sub-block to be processed;
and coding the bit width of the binary number corresponding to the maximum value, and coding each prediction residual in the sub-block to be processed according to the bit width.
6. An apparatus for compressing a reference frame, comprising:
a dividing module, configured to divide a reference frame data block into a plurality of compression units;
the compression module is used for compressing each compression unit;
the reference frame data block comprises a luminance data block and a chrominance data block, and is divided in the following way: taking each row of the brightness data block as a target row, and splicing every two rows of the chrominance data block as a target row to obtain a plurality of target rows; at least one target line in the plurality of target lines is taken as a compression unit;
the compression module comprises a prediction module for performing, for each of the compression units, the steps of:
adopting a 1eft prediction mode for each row of chrominance data in the compression unit; for the first line of brightness data in the compression unit, the first brightness data is not predicted, and other brightness data adopt a 1eft prediction mode; respectively predicting any remaining line of luminance data except the first line of luminance data in the compression unit by adopting an up prediction mode for the first luminance data in the any line of luminance data and respectively predicting other luminance data except the first luminance data in the any line of luminance data as a whole by adopting a 1eft prediction mode, an up-1eft prediction mode and an up-right prediction mode to obtain different prediction residual data blocks of the compression unit;
the apparatus is configured to perform the steps of:
and selecting one prediction residual data block from the prediction residual data blocks different from the compression unit for transmission, and coding and transmitting each prediction mode corresponding to the brightness data of any remaining row except the first row corresponding to the selected prediction residual data block.
7. The apparatus of claim 6, further comprising:
the memory dividing module is used for determining a rule of a storage unit with a fixed size according to the size of the storage space occupied by the reference frame data block, and dividing the storage space of an external memory into a plurality of storage units;
and a storage module, configured to store each total data stream into the external memory according to a rule that a total data stream obtained after one reference frame data block is compressed occupies one storage unit, where one total data stream is a sum of data streams obtained after compression units of one reference frame data block are compressed.
8. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 5.
9. A chip, comprising:
the computer-readable storage medium recited in claim 8; and
one or more processors to execute the program in the computer-readable storage medium.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710295126.XA CN108810556B (en) | 2017-04-28 | 2017-04-28 | Method, device and chip for compressing reference frame |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710295126.XA CN108810556B (en) | 2017-04-28 | 2017-04-28 | Method, device and chip for compressing reference frame |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108810556A CN108810556A (en) | 2018-11-13 |
| CN108810556B true CN108810556B (en) | 2021-12-24 |
Family
ID=64069171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710295126.XA Active CN108810556B (en) | 2017-04-28 | 2017-04-28 | Method, device and chip for compressing reference frame |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108810556B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101971633A (en) * | 2008-02-08 | 2011-02-09 | 线性代数技术有限公司 | A video coding system with reference frame compression |
| CN104885471A (en) * | 2013-01-30 | 2015-09-02 | 英特尔公司 | Content adaptive entropy coding of coded/not-coded data for next generation video |
| CN106375762A (en) * | 2015-07-22 | 2017-02-01 | 杭州海康威视数字技术股份有限公司 | Reference frame data compression method and apparatus |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101159874A (en) * | 2004-09-08 | 2008-04-09 | 华为技术有限公司 | intraframe coding based image chroma prediction technique |
| CN101087427B (en) * | 2006-06-06 | 2011-04-06 | 北京大学深圳研究生院 | A H.264 standard in-frame prediction mode selection method |
| EP2056606A1 (en) * | 2006-07-28 | 2009-05-06 | Kabushiki Kaisha Toshiba | Image encoding and decoding method and apparatus |
| CN101345874A (en) * | 2008-08-22 | 2009-01-14 | 北京中星微电子有限公司 | Video decoding method, reference frame processing method and apparatus |
| CN101394565B (en) * | 2008-10-20 | 2011-08-17 | 成都九洲电子信息系统有限责任公司 | Intra-frame prediction method |
| CN101820546A (en) * | 2009-02-27 | 2010-09-01 | 源见科技（苏州）有限公司 | Intra-frame prediction method |
| US8588536B2 (en) * | 2010-02-22 | 2013-11-19 | Texas Instruments Incorporated | Guaranteed-rate tiled image data compression |
| EP2617198A1 (en) * | 2011-01-12 | 2013-07-24 | Siemens Aktiengesellschaft | Compression and decompression of reference images in a video encoder |
| KR101307406B1 (en) * | 2011-08-05 | 2013-09-11 | 한양대학교 산학협력단 | Encoding/decoding apparatus with reference frame compression |
| CN103916664A (en) * | 2013-01-06 | 2014-07-09 | 北京大学 | Interframe mode selection device and method |
| CN105991951B (en) * | 2015-01-27 | 2019-03-08 | 炬芯（珠海）科技有限公司 | A kind of method and apparatus of video reference frame data storage |
| CN105245889B (en) * | 2015-09-24 | 2017-12-12 | 清华大学深圳研究生院 | A kind of reference frame compression method based on stratified sampling |
-
2017
- 2017-04-28 CN CN201710295126.XA patent/CN108810556B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101971633A (en) * | 2008-02-08 | 2011-02-09 | 线性代数技术有限公司 | A video coding system with reference frame compression |
| CN104885471A (en) * | 2013-01-30 | 2015-09-02 | 英特尔公司 | Content adaptive entropy coding of coded/not-coded data for next generation video |
| CN106375762A (en) * | 2015-07-22 | 2017-02-01 | 杭州海康威视数字技术股份有限公司 | Reference frame data compression method and apparatus |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108810556A (en) | 2018-11-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8422809B2 (en) | Image encoding method, image decoding method, image encoding apparatus, image decoding apparatus, image encoding program, and image decoding program | |
| CN113068032B (en) | Image encoding/decoding method, encoder, decoder, and storage medium | |
| US8767823B2 (en) | Method and apparatus for frame memory compression | |
| KR20220016232A (en) | Coding and decoding method, apparatus and storage medium | |
| JP6512928B2 (en) | Image coding apparatus, image processing apparatus, image coding method | |
| CN105828076A (en) | Flexible band offset mode in sample adaptive offset in HEVC | |
| CN109819250B (en) | Method and system for transforming multi-core full combination mode | |
| CN112740682A (en) | Scalar quantizer decision scheme for dependent scalar quantization | |
| JP6502753B2 (en) | Image coding apparatus, image processing apparatus, image coding method | |
| CN110944179B (en) | Video data decoding method and device | |
| WO2013089264A1 (en) | Image quantization apparatus, method and program, and image inverse quantization apparatus, method and program | |
| US10609382B2 (en) | Method and apparatus for compressing video data | |
| US11677932B2 (en) | Image processing device | |
| CN116489360A (en) | Image coding and decoding method and device | |
| CN108810556B (en) | Method, device and chip for compressing reference frame | |
| CN116095311A (en) | Encoding and decoding method, device and equipment thereof | |
| KR20100013142A (en) | Copression methode for frame memmory | |
| KR20210091657A (en) | Method of encoding and decoding image contents and system of transferring image contents | |
| US9560364B2 (en) | Encoding image data with quantizing and inverse-quantizing pixel values | |
| US20220046231A1 (en) | Video encoding/decoding method and device | |
| CN114567782B (en) | Raw image compression method and device suitable for 3DNR image noise reduction | |
| US11991367B2 (en) | Device and method for allocating code amounts to intra prediction modes | |
| CN116074529A (en) | Image encoding and decoding method, device and storage medium | |
| TW202348026A (en) | Image coding and decoding method and device, electronic equipment and storage medium | |
| CN116546193A (en) | Image decoding method, encoding method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: Zone C, floor 1, plant 1, No.1, Keji 4th Road, Tangjiawan Town, high tech Zone, Zhuhai City, Guangdong Province 519085 Applicant after: ACTIONS TECHNOLOGY Co.,Ltd. Address before: 519085 Guangdong province Zhuhai high tech Zone Science and technology innovation Coastal Science and technology No. four Road No. 1, No. 1 workshop C area Applicant before: ACTIONS (ZHUHAI) TECHNOLOGY Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |