WO2005122590A1 - 画像符号化装置および画像復号化装置、ならびにそれらで用いられる集積回路 - Google Patents
画像符号化装置および画像復号化装置、ならびにそれらで用いられる集積回路 Download PDFInfo
- Publication number
- WO2005122590A1 WO2005122590A1 PCT/JP2005/010419 JP2005010419W WO2005122590A1 WO 2005122590 A1 WO2005122590 A1 WO 2005122590A1 JP 2005010419 W JP2005010419 W JP 2005010419W WO 2005122590 A1 WO2005122590 A1 WO 2005122590A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- arithmetic
- data
- unit
- arithmetic unit
- conversion
- Prior art date
Links
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
- H04N19/42—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation
-
- 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
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/12—Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
- H04N19/122—Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
-
- 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
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
Definitions
- an image and a device and a product circuit used therewith, which relates to an image and a device having characteristics in conversion and conversion between a time domain and a frequency domain, and a product circuit used therewith.
- P2 has already been used in digital data, etc.
- P4 has been used in fields such as Internet mobile communication.
- the formula for the compression ratio was standardized by J (o deo ea). This is known as 264 PG4C or P4pa0 (shown below as 264).
- the book consisting of 6X6 elements of the code is divided into 4X4 books.
- the obtained 4X4 book is frequency-converted from the time domain by a bit and subtraction, placing it in the first place.
- the 4X4 book obtained in the frequency domain is converted from the frequency domain by bit and subtraction.
- the conversion for converting the time domain to the wave number domain by the bit and the subtraction is performed by the integer conversion.
- a conversion that converts a frequency domain into a time domain by subtraction is called an inversion ().
- Numeral 30 is a diagram showing the exchange of surface data at a66.
- 6X6 in the time domain is subjected to integer conversion in the 4X4 order, and converted to the frequency domain.
- four columns formed by extracting the C component contained in the 4X4 position are subjected to reduction by exchange.
- the function of a 6X 6 is realized.
- 00063 is a diagram showing the order of in 264.
- step S90 a66 it is determined that it is determined that the data is coded
- the data is written in a6d, and that it is not, the data to be coded is C minutes. Is replaced (Step S902), and the converted C component and other components are reversed.
- the encoded C component is converted (step S904), and the converted C component and other components are inverted (step 5905). If the data being processed is other data, the process is reversed.
- the inverted data is converted and obtained (step 907).
- 007 is the value of T C c, C for only C in step S90204.
- the component f which is obtained as a result, is the time domain obtained by the conversion.
- () means that is buffed rightward ().
- () means to bite leftward ().
- the numeral 3 indicates a value for changing the processing book size to 8 ⁇ 8 (42 in (2), (20) and (22)).
- Tw () (, 0,..., 7) is Indicates the area.
- T is a transmutation column.
- TX indicates the area of. According to FIG. 3, the range of can be obtained.
- 00115 indicates an equation for performing the conversion using the matrix a shown in 4. Five Here, indicates that you have selected 0 and ⁇ , 7. Then, according to the formula of 5, a0 to a7 and b0 to b7 can be calculated. Using b 0 to b 7, o 0 to o 7 for any are calculated. O 0 to o 7 obtained here are the values of the matrix wa for any given values. Next, the values of o 0 to o 7 are calculated, and the obtained values are set to 0 to 7. Then, similarly, the operation is performed by using 5 to obtain o0o7. Since the obtained o 0 o 7 is a part of the matrix a W a, it indicates the area of
- an object of the present invention is to provide an image device that can be downsized while complying with the standards.
- An arithmetic unit that performs arithmetic processing on a plurality of input data in an image device that performs compression data by switching the number of data from the frequency range to the frequency range as necessary. Select whether to read the operation data to save the operation result in the section or the data to be input to the operation section from the compressed data or to read from the data stored in the operation data.
- the order of input selection and input in which data is input to the operation unit corresponds to the number of operations that can be performed in one operation in the operation unit.
- the input selection controls the destination of only the data that enters the arithmetic section, and the arithmetic processing in the arithmetic section.
- An operation for controlling the number of multiplications in the arithmetic processing is provided.
- the order of the operations in the operation and the operation in the operation unit is defined by dividing the operation into an order that can be executed by the operation in the operation unit, the operation in each operation unit is executed by the single operation unit. can do. Therefore, an image device that can be downsized while complying with the standard will be provided.
- the arithmetic unit has a configuration capable of performing an arithmetic operation, controls only a combination of data to be calculated and calculated, and has a bit amount in the arithmetic unit. Is controlled as a multiplier.
- the arithmetic unit since the arithmetic unit has a configuration capable of performing the calculation, an image device that can perform the switching by performing the switching is provided.
- the operation unit includes a child operation unit of up to 8 children, a bit operation unit of the child of 2 children, a bit operation unit of 2 children that bites the power of the child of 3 children, and a bit operation unit of 3 children that bit the power of the child of 6 children.
- the bit unit the 4-bit unit that bites the power of the child of 7, the child's power
- the unit that adds the power of the second bit unit, and the power of the child's bit subtracts the power of the four children
- Arithmetic unit including the arithmetic unit of 2 that subtracts the power of the child of 5 from the power of the arithmetic unit of 4 and the arithmetic unit of 2 that adds the power of the child of 8 and the power of the third bit arithmetic unit , Operation data, and the arithmetic result of 2 and the arithmetic result of the arithmetic operation of 2 are stored, and based on the operation, conversion, and number, the data to be input to the children of up to 8 is controlled. Controls the amount of bits in the bit units of,.
- the plurality of keys are at least two keys of 4X4 conversion, 8X8 conversion, and conversion. 002 This makes it possible to realize the function of the H.264 in a small space.
- the order of operation and derivation is defined as being divided into operation and operation processing serving as a basic unit.
- the operation unit is configured to execute the operation processing serving as the basic unit. In order to repeat the arithmetic and basic arithmetic processing in the arithmetic unit, it controls only the data to be subjected to the arithmetic processing and controls the multiplication number.
- the basic arithmetic processing is performed in each mode.
- the arithmetic unit can be configured with a simpler configuration, and an image device that can be downsized while complying with standards can be provided.
- the basic arithmetic processing is an arithmetic processing capable of forming 22-power data.
- the data and number of operations and new conversions can be stored in correspondence with the number of operations, and when a new conversion is used, input selection and I will control the arithmetic part.
- the arithmetic unit performs a multiplication process on the powers of the children of to 4 and the child of the arithmetic unit.
- the arithmetic unit performs a multiplication process on the power of the child of 3, the arithmetic unit of 2, and the power of the child of 2.
- the arithmetic unit of 3 which performs multiplication processing on the arithmetic unit
- the arithmetic unit of 4 which performs multiplication processing on the power of the child of 4
- the power of the arithmetic unit includes the arithmetic unit of 2 to add the power of the fourth arithmetic unit.
- the result is stored and the data is controlled by the input selector when entering the children of ⁇ 4 based on the operation, conversion and number, and the multiplier which is the number in the arithmetic part of ⁇ 4.
- a multiplier that sets the number in the arithmetic unit of up to 4, a conversion unit that manages the conversion, and an arithmetic operation unit that manages the order in the arithmetic unit.
- the appropriate multiplier is read out from the multiplier section based on the operation number managed by the section, and set to the arithmetic section of up to 4, and the arithmetic operation section enters the children of up to 4 based on the current do and number. I will control the input data to input selection.
- a new conversion code may be used to store the multiplier set by the multiplier for setting the multiplier to be used in the new conversion function in accordance with the operation number.
- the multiplier and the multiplier corresponding to the operation number are read from the new unit and set in the arithmetic unit of up to 4, and the arithmetic operation unit enters the children of up to 4 based on the operation number. I'll control it.
- a plurality of keys are at least two of the following: sign conversion, 4X4 conversion, 8X8 conversion, and conversion.
- JP GPG 2 PG 4 PG 4 C 264
- Tana conversion An image device capable of responding to the password is provided.
- a plurality of arithmetic units may be connected in series or in series, and a plurality of arithmetic units connected in parallel or in series may be operated.
- the pipeline processing can be performed, and the processing degree can be improved.
- An image device for performing image processing by switching the number of times from the time domain to the frequency domain as necessary, and performing arithmetic processing on a plurality of input data.
- Calculation unit Select the operation to save the result, input to the operation unit, read out from the compression, or read out after it is stored in the operation, and input the read data to the operation unit.
- the order of input and output is defined according to the number of operations that can be executed in the operation unit in the operation unit, and the input selection enters the operation unit based on the conversion mode used and the number of operations in the operation unit.
- the arithmetic unit is provided with an operation for controlling only the data to be subjected to the arithmetic processing and the number of multiplications in the arithmetic processing.
- the arithmetic unit has a configuration capable of performing an arithmetic operation, controls only a combination of data to be calculated and calculated, and has a bit amount in the arithmetic unit. Is controlled as a multiplier.
- the operation unit is a bit operation unit for bidding the power of the child of up to 8 and the child of 2; a bit operation unit of 2 for bitting the power of the child of 3;
- the bit arithmetic unit, the bit arithmetic unit of 4 that bites the power of the child of 7, the bit arithmetic unit of 5 that bites the power of the child, the bit arithmetic unit of 6 that bites the power of the child of 4, and the child bit of the child of 5 The 7-bit arithmetic unit that bites the power, the 8-bit arithmetic unit that bites the power of the child of 8, the power of the 5-bit arithmetic unit
- the arithmetic unit that adds the power of the second bit arithmetic unit, and the bit arithmetic unit The arithmetic unit that subtracts the power of the 6-bit arithmetic unit from the power of the 4-bit arithmetic unit, the arithmetic unit that subtracts
- the order of operation and derivation is defined by dividing the operation into operation and basic operation processing.
- the operation unit has a configuration capable of executing the basic operation processing. In order to repeat the arithmetic and the basic arithmetic processing in the arithmetic unit, it controls only the data to be subjected to the arithmetic processing and controls the number of multiplications.
- the basic arithmetic processing is an arithmetic processing capable of forming 22-power data.
- the number of operations and the number of data associated with the new conversion mode can be stored in correspondence with the number of operations, and when a new conversion mode is used, the input selection and the calculation unit can be performed based on the contents.
- the arithmetic unit performs a multiplication process on the powers of the children of 4 and, the arithmetic unit which performs multiplication processing on the power of the child of 2, the arithmetic unit of 2 and the power of the child of ⁇ The arithmetic unit of 3 that performs multiplication on the arithmetic unit, the arithmetic unit of 4 that applies the multiplication process to the power of the child of 4, the arithmetic unit of the arithmetic unit, the arithmetic unit that adds the power of the second arithmetic unit, and the arithmetic unit of 3 Including the arithmetic part of 2 that adds the power of the fourth arithmetic part, saves the operation, and the result of the arithmetic operation of the two, and enters the children of up to 4 based on the operation, conversion and number.
- the data is controlled by the input selection, and the multiplier, which is the number in the arithmetic unit of, to be
- a multiplier for storing a multiplier corresponding to an operation and a conversion
- a multiplier that sets the number in the arithmetic unit of up to 4 a conversion unit that manages the conversion, and an arithmetic operation unit that manages the order in the arithmetic unit.
- the appropriate multiplier is read out from the multiplier section based on the operation number managed by the section, and set to the arithmetic section of up to 4, and the arithmetic operation section enters the children of up to 4 based on the current do and number. Control the input data to input selection.
- a new conversion function is used to store the multiplier set by the multiplier for setting the multiplier to be used in the new conversion function in accordance with the operation number.
- the multiplier and the multiplier corresponding to the operation number are read from the new unit and set in the arithmetic unit of up to 4, and the arithmetic operation unit inputs data to be input to the children of up to 4 based on the operation number Control to select.
- the plurality of keys are at least two of the following: sine, 4x4, 8x8, discrete sine, 4x4, 8x8, and exchange.
- a transducer that can handle all of these codes is provided.
- a plurality of arithmetic units are connected in series or in series, and the arithmetic unit controls a plurality of arithmetic units connected in parallel or in series.
- the operation unit includes a combination of data to be subjected to the operation processing and an operation to control a multiplication number in the operation processing.
- the data to be input to the arithmetic unit and arithmetic unit for processing is read from the compressed data, or the arithmetic unit Select the input that is read from the calculation result of, select the input to be input to the operation unit, and the order of the input and output in accordance with the number of operations that can be executed in one operation in the operation unit.
- the input selection controls the destination of data input to the operation unit based on the number of operations in the operation unit and the operation unit. And an operation for controlling the number of multiplications.
- Reference numeral 22 denotes a book showing an example of the generation of the image 2 according to the light state.
- 33 is a diagram showing a configuration of a converter 000 according to the light state.
- 44 is a diagram showing an arithmetic expression corresponding to the number of operations, and the relationship between the input to the human input and the input of the input process 02 and the operation processing.
- 55 is a char that indicates the operation of 0 when performing the calculation in the row direction in the 4X4 transformation.
- Reference numeral 6 denotes a char that indicates the operation of 0 when performing the direction calculation in the 4X4 conversion.
- 77 is a diagram showing an arithmetic expression corresponding to the number of operations, and a relation between data input to human power P to 8 by input selection 02. 8 8 is a chart showing the operation of 0 when performing a row-wise operation in 8X8 conversion.
- 999 is a diagram showing an arithmetic expression corresponding to the number of operations in the 4X4 conversion, and a relationship between data input to the inputs P to 8 by the input selection 02.
- 10A] 0 is a diagram showing an arithmetic expression corresponding to the number of operations in 8X8 conversion, and a relationship between data input to inputs P to 8 by input selection 02.
- 0B] B is a block indicating the formation of the converter 000a capable of 8X8 conversion.
- FIG. 11 is a diagram showing a configuration of a converter 2000 according to the second embodiment.
- 122 is a block diagram showing the configuration of the arithmetic unit 00 of FIG.
- 133 is a block diagram showing the configuration of the arithmetic unit 200 of 2.
- 144 is a block diagram showing a configuration of the arithmetic unit when implementing the 8X8 conversion by connecting the arithmetic units in series and / or in parallel.
- 155 is a diagram showing a formula in the arithmetic unit 00 and a formula in the arithmetic unit 200 of 2 corresponding to the number of operations.
- 16 6 is used to perform row-wise arithmetic in 8X8 conversion.
- 177 is a block diagram showing a schematic configuration of image 3.
- 88 is a block diagram showing a schematic configuration of the image 4.
- 199 is a block diagram showing the formation of the conversion 3000.
- Reference numeral 20 20 denotes data for explaining the flow of the operation in the operation unit 300.
- 22 is a diagram showing C-commutation data.
- 22 22 is a diagram showing an 8 ⁇ 8 conversion duff.
- 23 23 is a view showing a 8 ⁇ 8 conversion duff.
- 24 24 is a diagram showing data of the 4X4 conversion.
- 25 25 is a diagram showing 4X4 conversion data.
- 26 26 is a diagram showing 4X4 conversion data.
- 27 27 is a diagram showing a 2 ⁇ 2 conversion duff.
- 28 28 is a block showing the composition of the converter 4000 according to the state of 4 29
- 29 is a block showing the composition of the converter 5000 according to the state of 5
- 30 is a 6 ⁇ 6 313, which is a diagram showing the exchange of data, is a diagram showing the order of ⁇ ⁇ ⁇ ⁇ in 264.
- 300 data 3002 Power selection 3003 arithmetic unit 3004 arithmetic unit 3005 arithmetic unit 3006 arithmetic unit 3007 arithmetic unit 3008 arithmetic unit 3909 data 30 0 arithmetic unit 30
- FIG. 0056 is a block diagram showing an example of the generation of an image according to the light state. Here, the image is displayed on the screen,
- the calculation unit 7 adds the force from the motion compensation 20 in the screen or from the inverse 6.
- the booking filter 8 removes the book noise of the force from the adder 7.
- Fume 9 stores the fume.
- the compensation 20 performs motion compensation, predicts weighting, and outputs the hood multiplied by the number of weights to the subtraction unit 22.
- the arithmetic unit 22 subtracts the hum from the motion compensation 20 for the hum outside one picture, and sets the result to 2. 2 detects motion and makes it 4.
- the amp 4 executes the amp based on the query. This has improved motion compensation
- 00052 is a block diagram showing an example of the formation of image 2 according to the light state. 2
- the image 2 includes the following symbols 23, 24, 25, a booking iterator 26, a motion compensation 27, a weight prediction 28, an in-screen 29, and an adder 30.
- the 006 Book filter 26 removes the power block noise from 25.
- the compensation 27 performs motion compensation based on the force from the amp 23.
- the weighting prediction 28 predicts the weight as a result of the motion compensation, and calculates the hum that is multiplied by the weight. 29 in the plane Based on these forces, it predicts the difference between the units, and controls the switching between the adding unit 30, the in-screen 29, and the weighting prediction 28. As a result, a video signal with improved motion compensation is output.
- the image in the embodiment has a number of numbers from the time domain to the frequency domain, and switches the number as needed to form an image. Further, the image according to the present embodiment has a number of numbers from the frequency range to the range, and the image is restored while switching the number as necessary.
- 0063 is applicable to configurations and configurations other than the configurations disclosed in the above and 2, which are disclosed and replaced below.
- the present invention can be applied to the cases other than the above.
- 00643 is a diagram showing the configuration of the converter 000 according to the light state.
- the converter 000 applies to 2 and 6 shown in FIG. And, the replacement 000 shown in 3 applies even if it is outside or in the location, or if it is replaced.
- the converter 000 includes an operation unit 00, an operation 0, an input selection 02, data, and an operation 2.
- an operation unit 00 the converter 000 includes an operation unit 00, an operation 0, an input selection 02, data, and an operation 2.
- the arithmetic unit 00 performs arithmetic processing on a plurality of input data.
- the arithmetic unit 00 has ⁇ 8, a bit arithmetic unit 03 304 506, an addition unit 07 70, and a subtraction unit 08 09.
- the bit unit 04 bit off the power of 2.
- the bit unit 03 turns off the power of p3.
- the bit unit 06 bit off the power of p6.
- the bit operation unit 05 bit-forces the power of.
- the arithmetic unit 07 adds the power of P and the power of the arithmetic unit 03.
- the arithmetic unit 08 subtracts the power of 4 from the power of the bit arithmetic unit 04.
- Calculation part 09 Subtracts the power of 5 from the power of bit unit 05.
- the arithmetic unit 0 adds the power of 8 and the power of the arithmetic unit 06.
- the arithmetic unit 00 has a configuration capable of performing the arithmetic. 2 stores the calculation result of the addition unit 07 0 and the calculation unit 08 09.
- the input selection 02 selects the current number of row data to be read from the data, read from the data in the row direction, or read from the calculation data 2 in the row direction. Do it.
- the operation 0 determines the bit amount of the bit calculation unit 03 04 05 06 depending on whether the current and the logic are changed, 4 ⁇ 4 conversion, or 8 ⁇ 8 conversion, and calculates the bit calculation unit 03 304 05 , 06 is issued.
- 0 is defined in the order of “do” in a place where it can be executed in the arithmetic operation of the arithmetic unit 00 and corresponding to the number of operations. 0 controls the output destination of data only when the input selection 02 is supplied to the operation unit 00 based on the number of conversion fields and the number to be used, and the data to be subjected to operation processing in the operation unit 00. Controls multiplication (bits) in combination and arithmetic processing.
- the force selection 02 is based on the force selection signal from the operation 0 and the like.
- the input selection 02 recognizes only the data to be subjected to the arithmetic processing based on the number indicated by the force selection signal, and changes the read data from an appropriate P to 8.
- bit arithmetic unit 03 04 05 06 performs either the no bit, the bit, or the two bit data for the data input from the input selection 02 according to the bit number input from the operation 0. Bit addition data is added to the adder part 07 0 and the subtractor part 08 09. Note that the data output from the bit arithmetic units 03, 04, 05, 06, Data rub.
- the 007 arithmetic unit 0770 calculates the bit data from the bit arithmetic unit 03306 and the value input from P8.
- the arithmetic unit 08 09 subtracts the data input from p4 and p5 from the bit data from the bit arithmetic unit 04 05.
- Data is a notation such as "me”, and stores data to be replaced.
- Data 2 is a notation such as a character, and stores the data converted by the arithmetic unit 00. In addition, the arithmetic data 2 performs the row data readout and the row write readout in order to perform the data calculation.
- the matrix data for example, W
- the matrix for conversion for example, W
- the obtained square matrix W is multiplied by the square matrix from the left.
- the square matrix W obtained in this way is the transformation data.
- an operation of multiplying a square matrix W by a square row from the right is a row-wise operation.
- an operation of multiplying the square matrix W from the left is a direction operation.
- FIG. 4 is a diagram showing an arithmetic expression corresponding to the number of operations, and a relation of only data to be input to inputs P to p8 by input selection 02 and to be subjected to arithmetic processing. As shown in Fig. 4, the relation is defined as operation 0. However, the present invention is not limited to the table type as in 4, but is not particularly limited.
- the calculation of the row direction and the direction is shown.
- the data to be input is represented by the 4X4 columns.
- the input data is The component of the row with 4X4 column is reduced to 0-3.
- 0 is input into 2
- 2 is input into 3 and 4
- 2 is input into 7 and 8
- 3 is input into 5 and 6.
- a0 to a3 are obtained.
- a0 to a3 are input to the corresponding ⁇ 8.
- b0 to b3 are obtained.
- Two operations are performed over the rows of the input 4X4 surface data. As a result, the calculation in the row direction is completed, and 4X4 plane data is obtained.
- the direction calculation is performed on the 4X4 data column obtained by the row direction calculation.
- the elements of a certain column of the matrix obtained by km 0 to 3 are calculated.
- 4 to 3 are input to P to 8.
- a0 to a3 are obtained.
- a0 to a3 are input to the corresponding .about.8.
- b0 to b3 are obtained.
- Two operations are performed over all 4X4 columns. This completes the calculation of the direction.
- 4X4 converted into 4X4 is obtained.
- 00805 is a char- acter indicating the operation of "0" when performing the arithmetic operation in the row direction in the 4X4 conversion.
- FIG. 5 referring to FIG. 5, in the 4X4 conversion, a description will be given of the operation of “0” when performing the calculation in the row direction.
- Step 2 ends when the number in the row direction shown in 4 reaches two.
- input matrix of 4X4 data input matrix of 4X4 data.
- step S In the calculation of the direction of the matrix (step 2, step S), operation 0 outputs an input selection signal and makes input selection 02.
- the force selection signal includes that the number of operations for reading the row components in the input matrix from the input data is an eye (step S).
- the input selection 02 In response to the power selection signal, the input selection 02 reads out the data of 4 from the data and makes the terminals P to 8 as shown in 3. Next, operation 0 outputs a bit. Bit
- the signal includes an instruction to the bit arithmetic unit 03 04 that determines that there is no bit, and an instruction to the bit arithmetic unit 05 06 to the right (Step S 02) 0085 According to the bit number, the bit arithmetic unit 05 06 Generates bit-by-bit data from the input data, and performs
- the arithmetic unit 07 adds the power P of the bit arithmetic unit 03.
- the arithmetic unit 08 subtracts the power of 4 from the power of the bit arithmetic unit 04.
- the arithmetic unit 09 subtracts the power of p5 from the power of the bit arithmetic unit 05.
- the arithmetic unit 0 adds the power of the bit arithmetic unit 06 and the power of 8. The result obtained by these subtractions is the result of the calculation in the row direction.
- Data 2 stores the calculation result in the row direction of f days.
- step 22 step S 002
- the operation 0 outputs an input selection signal and makes the input selection 02.
- the calculation result in the above direction is calculated
- Step S03 which includes the fact that the number of operations to read from 2 is the second.
- the input selection 02 extracts the row direction calculation result stored in the operation data 2 from the row direction, and
- operation 0 outputs a bit signal.
- the signal includes an instruction to the bit calculation unit 03 04 05 06 to make no bit (step S 04).
- the bit calculation unit 03 04 05 06 performs the data input without performing the bit conversion.
- the operation unit 07 0 and the operation units 08 and 09 execute the operation based on the input data, and store the operation result of the second row direction in the operation data 2 in the fth position. This ends the process of day 2.
- step 2 The remaining three lines of the process of step 2 are performed. This satisfies the condition of the project. As a result, 4 is obtained by the calculation in the row direction. Data 2 stores the data of 4 as the result of the calculation in the T direction.
- Reference numeral "00936” denotes a feature indicating the operation of "0" when the direction calculation is performed in the 4X4 conversion. In the following, with reference to FIG. 6, a description will be given of the operation of calculating a zero in a 4X4 conversion.
- step 3 in the step (4) of the calculation result in the T direction obtained by the calculation in the row direction, the processing is completed when the operation in the step is completed.
- Step 4 ends when the number of directions shown in 4 reaches two.
- the operation 0 outputs the input selection signal and makes the input selection 02.
- Force selection code is stored in operation data 2.
- the input selection 02 reads out the T-direction calculation result stored in the operation data 2 in the direction, and performs 4 to 3 to 3 to the terminal to 8 as shown in 3.
- the signal includes an instruction to the bit arithmetic unit 03 04 to be set to no bit and an instruction to the bit arithmetic unit 05 06 to the right (Step S202) 0098
- the bit arithmetic unit 05 06 generates input data and generates bit data.
- the calculation unit 07 adds the power P of the bit calculation unit 03 to the power.
- the arithmetic unit 08 subtracts the power of 4 from the power of the bit arithmetic unit 04.
- the arithmetic part 09 is Subtract 5 power from the power of arithmetic unit 05.
- the arithmetic unit 0 adds the power of the power 8 of the bit arithmetic unit 06.
- the result obtained by these subtractions is the result of the eye. 2 saves the result in the direction of.
- operation 0 In the calculation of 2 (step 4200, step S2002) of the calculation result, operation 0 outputs an input selection signal and makes input selection 02.
- the force selection signal includes the fact that the number of operations is the second to read the result of the above operation from operation data 2 (step S203).
- the input selection 02 reads out the calculation result stored in the calculation data 2 from the direction and outputs the data of 4 to the P corresponding to the second calculation as shown in 3.
- operation 0 outputs a bit. Bit
- the signal contains an instruction for the bit arithmetic unit 03 04 05 06 to be removed.
- the bit calculation unit 03 04 05 06 is input without performing any bit.
- the arithmetic unit 0770 and the arithmetic unit 0809 perform the operation based on the input data, and store the second calculation result in the operation data 2 in the row direction. This ends the process of step 4.
- 0150 0 has three lines remaining in step 4. This allows
- Condition 3 is satisfied.
- 4 data can be obtained by column-wise calculation.
- 4 is the final result obtained by the 4X4 transformation. Therefore, the operation data 2 is obtained by using the data of 4 as an output matrix.
- converter 000 can perform 4X4 conversion.
- FIG. 9 is a diagram showing a relationship when input is made to (8). As shown in Fig. 7, the relation is defined in operation 0. , 7 is not limited to the table type, but is not particularly limited.
- the input data is represented by 8X8 columns. In the case of 7, the input data is included in the component and there is an 8X8 column), and 0 to 7 are set. Only the data input to P to p8 are supported in the calculation of the 8th to 8th. By the calculations of the 8th to 8th, the operation results defined by the operation expressions are obtained. The calculation in the row direction () is completed by the calculation up to the 8th shown in Fig. 7, and 8X8 plane data is obtained.
- a reference numeral “01098” denotes a feature indicating the operation of “0” when performing an arithmetic operation in the row direction in the 8 ⁇ 8 conversion. Referring to FIG. 8 below, in the 8X8 conversion, a description of the operation of 0 when performing the arithmetic in the row direction is performed. To complete when the operation in step 5 is completed. Step 6 ends when the number in the row direction shown in 8 reaches eight. Below, the input 8x8 input matrix.
- the operation 0 In the calculation of the direction of the matrix (step 6, step S300), the operation 0 outputs the input selection signal and makes the input selection 02.
- the force selection signal includes the fact that the number of operations for reading 426, which is the value of f in the input matrix, from the data is an eye (step S30).
- bit arithmetic unit 03 04 In response to the force selection signal, input selection 02 reads inputs 0, 2, 4, 6 from the data, and from P to p8, as indicated by 7. [0114] Next, operation 0 outputs a bit.
- the bits include an instruction to the bit arithmetic unit 03 04 to be removed and an instruction to the bit arithmetic unit 05 06 to the right (Step S302). 0115 In response to the bit number, the bit arithmetic unit 05 06 Generates biff-biff data based on the input data.
- the arithmetic unit 07 adds the power P of the bit arithmetic unit 03.
- the arithmetic unit 08 subtracts the power of 4 from the power of the bit arithmetic unit 04.
- the arithmetic unit 09 subtracts the power of p5 from the power of the bit arithmetic unit 05.
- the arithmetic unit 0 adds the power of the bit arithmetic unit 06 and the power of 8. The result obtained by these subtractions is stored in the operation data 2 in the row direction as the row direction calculation result.
- operation 0 In the calculation of direction 2 (step 62, step 3002), operation 0 outputs an input selection signal and makes input selection 02.
- the force selection signal includes that the operation number is the second for reading the row direction calculation result from the operation data 2 (step 5303). 0118 According to the force selection signal, the input selection 02
- the row direction calculation result stored in the calculation data 2 is read, and the data of 4 is set to 8 as shown in 7 if it corresponds to the second calculation.
- the signal includes an instruction to the bit arithmetic unit 03 04 05 06 to be bit-less (Step S304)
- the bit calculation unit 03 04 05 06 performs the input data without bit-off.
- the 012 arithmetic unit 0770 and the arithmetic unit 0809 execute the operation based on the input data, and store the operation result in the row direction 2 in the operation data 2.
- operation 0 is the third and fourth operation that is performed only from data. S30 to output the input selection number.
- the operation 0 instructs the bit operation unit 03 304 to not perform the bit operation, and outputs a bit indicating the bit operation to the bit operation unit 0506.
- Input selection 02 reads out the data 357 from the data, and changes from 8 to 8 as in 7 according to the number of operations.
- the bit arithmetic unit 05 06 shifts the data to the right.
- the arithmetic unit 07 0 and the arithmetic unit 08 09 perform operations. data
- the operation 0 calculates the row direction directional calculation result and the direction directional operation result in the operation data.
- the operation 0 outputs a bit instructing the bit units 03, 04, 05, and 06 not to perform a bit operation (S304).
- the input selection 02 reads the row directional calculation result and the directional calculation result stored in the data, and sets the data of 8 to 8 as corresponding to 7 as in 7.
- the bit calculation unit 03 04 05 06 does not shift the input data.
- the arithmetic unit 07 0 and the arithmetic unit 08 09 perform operations. data
- operation 0 In the calculation of direction 6 (step 66, step 3006), operation 0 outputs an input selection signal indicating that it is the sixth operation in order to read the result of operation 5 in the row direction from operation data 2.
- the operation 0 is performed on the bit units 03 04 05 06
- the input selection 02 reads the row direction math result stored in the arithmetic data 2 and outputs the result as shown in 7. , Corresponding to p 8.
- the bit calculation unit 03 04 05 06 converts the input data into 2-bit data and bit-ft data.
- the arithmetic unit 07 0 and the arithmetic unit 08 09 perform operations. data
- step 607, step 3007) and direction 8 step 608, step S3008
- operation 0 is read from the row direction 2 result row direction 6 calculation result operation data 2
- this is the calculation of the 7th and 8th operations, and the input selection signal is issued.
- the operation 0 outputs a bit instructing the bit operation unit 03 304 05 06 not to perform a bit operation (Step S304).
- the input selection 02 reads the row direction 2 calculation result stored in the operation data 2 and the row direction 6 calculation result, and converts the data of 4 to 8 as shown in 7 when the corresponding value is obtained. I do.
- the bit calculation unit 03 04 05 06 performs the input data bit-by-bit.
- the arithmetic unit 07 0 and the arithmetic unit 08 09 perform operations. data
- step 6 The process of step 6 is the remaining seven lines. As a result, the direction calculation result is obtained and stored in calculation data 2.
- the converter 000 calculates the obtained result in the T direction, Perform the calculation of.
- the operation 0 performs the eight operations shown in step 6 of the result of the T-direction operation in the operation unit 00 to obtain column data, and obtains eight columns of data in total.
- the arithmetic unit 00 performs eight operations shown in steps 6 and 7.
- the eight-column image data obtained by the calculation in this direction is the final result obtained by the 8X8 conversion. Therefore, the operation data 2 uses 8 data as an output matrix.
- the converter 000 can perform the 8X8 conversion.
- the operation 0 is defined by dividing the order of the 4X4 and 8X8 conversions in such a way that the operation can be executed by the number of operations in the operation unit 00 according to the number of operations. Then, in accordance with the operation, the number of conversions to be used, and the number, the input selection 02 controls only the data to be set to P to p8, and controls the bits () in the bit calculation units 03 to 06. Thus, 4X4 and 8X8 conversion can be realized using a single converter 000.
- the input selection 02 changes the data to terminals to 8 as shown in 9.
- 0 is operated so that the bit amount in the bit calculation unit 0506 is shifted to the left. In this way, the 4X4 conversion can be realized by the route shown in 3.
- the input selection 02 sets the data to the terminals P to 8 as shown by 0. 0 adjusts the amount of biff and performs arithmetic control.
- bit units 03a 04a 05a 06a also on the side of p p4 p5 p8.
- f corresponds to, corresponds to a mill, and corresponds to o. It should be noted that, b, b b2, 7 and can be understood.
- C (0, ⁇ .. 3) indicates the extracted C component. 0 45 () expands to 9
- the converter 000 may execute the operation in the following order.
- Step (Step) 0 reads the r4 number of 4 ⁇ 4 (C) of C horizontally, and sets the corresponding value to ⁇ 8.
- which child to enter can be defined from the above expression.
- Step 2 0 outputs a bit signal indicating that no bit is output.
- Step 3 In the second calculation, the addition unit 07 0 and the calculation unit 08 09 calculate,, and C indicated by 0, and store them in the operation 2.
- Step 4 In the second calculation, calculation is performed by calculation 0
- Equation 2 is obtained.
- Step 6 the operation 0 performs the calculation of the direction to obtain the matrix shown in 3. Since the calculation in this direction is the same as the calculation in the direction up to step 5, it can be performed by the calculation unit 00.
- the operation 0 is divided into units that can be executed in each operation in the operation unit, and is defined corresponding to the number of operations. Therefore, according to the number of operations, the operation 0 can perform the conversion of a66 by controlling the amount of data input to the operation unit 00 and the amount of bits. . Note that, in the above description,
- the conversion is the same even in the case of, and therefore, the conversion can be performed using the converter 000.
- C (,) indicates the extracted C component.
- Step (Step) 0 reads the 4X of 2X2 (C) into the input selection 02 and sets it to the corresponding p8. It should be noted that which child to enter can be easily estimated from the above formula, and therefore will be omitted.
- Step 2 and C are output by the arithmetic unit 07 0 and the arithmetic unit 08 09 and stored in the arithmetic data 2.
- Step 3) 0 reads,, C, stored in the operation data 2 to the input selection 02 and inputs them to P to p8 so that the calculation of 9 can be performed.
- the arithmetic unit 00 can calculate the result shown in FIG. [0160]
- the operation 0 is divided according to the number of operations that can be executed by the operation in the operation unit, and is defined according to the number of operations. Therefore, according to the number of operations, the operation 0 can perform the Coa conversion by controlling only the amount of data and the bit amount in the operation unit 00. Note that, in the above description, the Coa conversion in the case of the code has been described. However, the conversion is the same in the Coa conversion in the case of, so that the conversion can be performed using the converter 000.
- operation 0 is performed in the order of 4X4 conversion, 8X8 conversion, 4X4 conversion, 8X8 conversion, and conversion.
- it is defined in such a way that it can be executed in the arithmetic operation of the arithmetic unit 00 in accordance with the number of operations.
- the input selection 02 is controlled to be 8 or less, and the bits () in the bit calculation units 03 to 06 are controlled.
- 4X4 conversion, 8X8 conversion, 4X4 conversion, 8X8 conversion, and switching can be realized using a single converter 000.
- the details are as follows. In the 8X8 conversion shown in Fig.
- the fifth equation is as follows: enter 4 (for example, aa 5 a2 a25), and perform 2 (for example, aa5). Is one set (), and thus there are two sets of arithmetic expressions (and 2). The other formulas are those that input a value of 2 and output a value of 2. Circuits that can implement the fifth formula in the 8X8 conversion include p to 4, the bit calculation unit 03 04, the addition unit 07, and the calculation unit 08. In the arithmetic unit 00, further low-order components are prepared (5 to 8, the bit arithmetic unit 05 and 06, the subtraction unit 09, and the arithmetic unit 0). The construction of the limit is based on a 42-power circuit.
- the 42-power circuit functions as a 22-power circuit if the same value is input to the second child, as in the formula outside the fifth formula in 8X8 conversion. Therefore, 4X4 conversion, 8X8 conversion, 4X4 conversion, 8X8 conversion, a66 (4X4), and Coa (2X2) can be calculated using a 42-power circuit or a 22-power circuit.
- control is performed by using and bit 0, conversion can be realized using a common converter.
- the image and the single converter can be used in a time-division manner while changing the standard to be used as necessary.
- Reference numeral 0166 is a block diagram showing a configuration of the converter 2000 according to the second embodiment.
- the converter 2000 according to the second embodiment is applied to 2 and 6 shown in (2) and 25 (2).
- the exchange 2000 shown in the following applies.
- the converter 2000 includes an arithmetic unit 00, an arithmetic unit 200 of 2, data, operation data 2, and operation 20.
- Reference numeral 0168 2 denotes a block indicating the formation of the arithmetic unit 00 of the equation. 2
- the arithmetic unit 00 has an input selector 02, a bit arithmetic unit 03 04 05 06, an addition unit 70, and a subtraction unit 08 09.
- the configuration of the arithmetic unit 00 is the same as that of the arithmetic unit 00 shown in FIG. 01163 is a block diagram showing the configuration of the arithmetic unit 200 of 2.
- the arithmetic unit 200 has an input selection unit 202, a bit arithmetic unit 203.04.05.06, an addition unit 2007.0, and a subtraction unit 208.09.
- the configuration of the arithmetic unit 200 of 2 is the same as that of the arithmetic unit 00 shown in FIG. 0170 20 is the power selection of the arithmetic part 00 of 02.
- the operation 20 determines the bit amount of the bit calculation unit 03 304 05 06 based on the number of conversions (4 conversion, 2X2 conversion, 4X4 conversion, 8X8 conversion, 4X4 conversion, or 8X8) operation. The determined bit shift amount is used as the first bit number for l 03 04 05 06. As in 017, the operation 20 is based on the power selection 202 of the arithmetic unit 200 of 2
- bit amount of the bit calculation unit 203 304 05 06 is determined, and the determined bit amount is set as the second bit number for the bit calculation unit 203 304 05 06.
- the operation 20 uses the arithmetic unit 00 to perform the conversion processing and output the power selection and the BIFF signal.
- the converter 2000 has the following configuration except that the arithmetic unit 200 of 2 is not used.
- 0173 20 is for 8X8 and 8X8 conversions
- the conversion degree is improved by using the arithmetic unit 00000 of 2.
- 5 is a diagram showing an equation in the arithmetic unit 00 and an arithmetic expression in the arithmetic unit 200 of 2 corresponding to the number of operations. As shown in 5, the relationship is defined in operation 20. However, the present invention is not limited to the table type as described in 5, but is not particularly limited.
- the arithmetic unit 00 200 of and 2 reads data necessary for the arithmetic corresponding to the processing number from the data or arithmetic data 2 and executes the same arithmetic.
- Reference numeral 01766 denotes a feature indicating the operation of 20 when performing the arithmetic operation in the row direction in the 8X8 conversion. Refer to Fig. 6 below for 8X8 conversion. Here, we explain about the work of 20 when performing the calculation in the row direction.
- the step 7 is a step that terminates the completion of the operation in the step 7 in the parentheses of the input 8 ⁇ 8. Step 8 ends when the number in the row direction shown in 6 reaches four.
- input matrix of input 8X8 plane data is a feature indicating the operation of 20 when performing the arithmetic operation in the row direction in the 8X8 conversion.
- the operation 20 In the calculation of the direction of the matrix (step 8, step S4), the operation 20 outputs the force selection signal of and inputs it to ⁇ 00 (step S4), and outputs the force indication signal of 22
- the calculation unit 200 is calculated (step 402).
- the force selection signal includes a notification that the processing signal is 3 for reading the input 357 from the data.
- the force selection signal of 2 includes a message indicating that the processing signal is 4 so that only input 357 is obtained from the data.
- the input selection 022 operates in the same way as the third and fourth calculations of the input selection 02 in. 0180 Next, operation 20 outputs the bit of
- the signal includes an instruction to bit the bit units 05 and 06.
- the bit 2 includes an instruction to bit the arithmetic unit 205506.
- the power selection signal of includes a message indicating that the processing number is 5 for reading out the result of the second row-direction operation stored in Operation 2.
- operation 20 outputs the bit number of (step 43).
- the bit of 2 is output (step 44).
- the bit number includes an instruction to bit off to the bit units 05 and 06.
- the bit number of 2 includes an instruction not to perform a bit operation on the bit calculation unit 203 04 05 06.
- the arithmetic unit 00 00 of and 2 operates similarly to the arithmetic operation of the arithmetic unit 00 and the fifth arithmetic unit in the following state.
- the output unit 20 outputs the power selection signal and inputs it to the arithmetic unit 00 (step S42), and outputs the power input signal of 2 to make it into the arithmetic unit 200 (step S422).
- the power selection signal includes a message indicating that the processing signal is 2 for reading out the calculation result of the second operation in the row direction from the operation data 2.
- the power selection signal of No. 2 includes a message indicating that the processing number is 6 for reading out the calculation result of the second operation in the row direction from the operation data 2.
- input selection 02 02 operates in the same way as the second and sixth arithmetic of input selection 02 in the following condition. [0188] Next, the operation 20 outputs the bit number of (step 423).
- a bit number of 2 is output (step 424).
- the bit number includes an instruction not to perform bit-off for the bit calculation unit 03 04 05 06.
- the 2 bit number includes an indication for the 2 bit for the bit unit 203.405.06.
- the arithmetic unit 00 00 of and 2 operates in the same manner as the arithmetic operation of the arithmetic unit 00 in the second and sixth arithmetic operations.
- the power selection signal includes a message indicating that the processing signal is 7 in order to read the third calculation result in the row direction from the operation data 2.
- the power selection signal of No. 2 includes a message indicating that the processing number is 8 in order to read the third calculation result in the row direction from the operation data 2.
- the input selection 022 operates in the same way as the calculation of the seventh and eighth inputs in the input selection 02. [0192] Next, operation 20 outputs the bits of
- the signal includes an instruction not to perform a bit operation on the bit operation unit 03 04 05 06.
- the 2 bits include an instruction not to perform a bit operation on the bit calculation section 203 03 05 06.
- the arithmetic unit 00 00 of and 2 operates in the same way as the arithmetic operation of the arithmetic unit 00 in the 7th and 8th operations.
- step 8 The process of step 8 is executed for the remaining seven lines. As a result, an 8X8X8T direction calculation result is obtained and stored in calculation data 2.
- the converter 2000 performs the direction calculation on the obtained T direction calculation result, and performs the four calculations shown in step 8 on the T direction calculation result in the direction calculation in the direction calculation.
- the eight-row surface data obtained by the calculation in this direction is the final result obtained by the 8X8 conversion. So, 2 outputs 8 data as an output matrix, and thus, the converter 2000 can perform 8X8 conversion.
- the operation is performed in parallel by connecting the two converters 00 00 and 00 00 in parallel.
- the arithmetic unit 00 shown in 3 is connected in series, the operation, necessary data is obtained by the power selection 02 of the arithmetic unit 00, and the following operation is performed while the arithmetic unit 00 of the subsequent stage performs the operation.
- the operation may be performed in the previous arithmetic unit 00.
- the fourth calculation is performed by the previous calculation unit 00
- the second calculation is performed by the calculation unit 200 in the subsequent stage
- the calculation unit 00 is sequentially processed in the row direction data.
- the result in the row direction can be sequentially obtained from the arithmetic unit 200 in the subsequent stage, and the result is improved.
- the converters are connected in parallel or in series, and if the input and the bit amount are controlled by the arithmetic operation, the converter can perform the conversion processing correctly.
- the operation unit shown in FIG. 4 can execute the operation corresponding to the number 7.
- the degree of 8X8 conversion can be improved.
- data will be output from the broken line by using only the second arithmetic part.
- the 4X4 conversion, 4X4 conversion, 4X4 conversion, and 2X2 conversion also May be connected in parallel or in series to perform pipeline processing.
- the processing degree is improved.
- the processing degree may be improved by using a different arithmetic unit.
- the operation 20 defines whether or not to execute a necessary operation by inputting necessary image data to a shift operation unit of a plurality of operation units in accordance with an operation number. Therefore, multiple arithmetic units are used in parallel or in series. In this way, by connecting the basic operation units in parallel or in series, it is possible to perform pipeline processing with a simple configuration, and to improve image processing and image processing.
- FIG. 7 is a block diagram showing a schematic configuration of the image 3. In FIG. 7, only the configuration for converting the time domain into the wave number domain and compressing it, and omitting the configuration for reducing it by motion compensation are omitted. 7
- Data 3 includes image data 3, time and wave numbers 32 and 33, and compressed data 34.
- Data 3 converts data in the time domain to a time wave number of 32.
- Wave number 32 converts data in the time domain into data in the wave number domain.
- conversion from the time domain to the frequency domain is performed by using any of the C4 conversion, the 8X8 conversion, the 4X4 conversion, the 2X2 conversion, and the newly added known conversion method.
- 33 changes the data in the wavenumber range
- the original image data is compressed.
- Data 34 outputs and saves a compressed image.
- 0258 is a block diagram showing a schematic configuration of the image 4.
- the image 4 includes compressed data 4, 42, frequency time 43, and image data 44.
- Data 4 uses compressed image data.
- Wave number time 43 converts wave number data into time domain data. Wave number time 43 is C4 conversion, 8X8 conversion, 4X4
- the data 44 outputs and saves the data in the time domain obtained by the frequency time 43.
- Wave number 32 and wave number time 43 are C, C 4X4 conversion, 4X4 conversion, 8X8 conversion, 8X8 conversion, 4X4
- Wave number 32 and wave number time 43 can be realized by a common composition. Below, time wavenumber 32 and wavenumber time 43 will simply be referred to as transform 300.
- Numeral 22979 is a block diagram showing the result of conversion 3000.
- the converter 3000 includes data 300, an input selection 3002, an operation unit 300 0, operation data 3009, a multiplication 30, an operation processing unit 30 2, and a conversion unit 303 4, 2 of 3
- the arithmetic unit 300 includes a multiplying unit 3003 004 005 006 and an adding unit 3007 008.
- 30 and the arithmetic processing unit 30 2 the conversion registers 30 3,
- the arithmetic operation is performed by the following formulas: 30 5, new de 30 6, and new 30 7.
- the conversion schemes that can be used in the form of 0283 are C, C 4X4 conversion, 4X4 conversion, 8X8 conversion, 8X8 conversion, 4X4 conversion, 2X2 conversion, and other new conversion methods to be added.
- the default value is set to the conversion value 303 by the new processor.
- 0209 data 300 is used to store data to be converted.
- the data 3009 stores the calculation data output from the calculation unit 3000.
- the 0210 processing unit 302 manages the number for conversion in accordance with the conversion data managed by the conversion table 303, and notifies the input selection 3002 whether it is the current calculation.
- the force selection unit 3002 reads out necessary data from the data 300 or the operation data 3009 according to the knowledge from the arithmetic processing unit 3002, and reads out the appropriate data from (4) to (4). Depending on the conversion method and number, input selection
- Numeral 3041 stores C and a multiplier (numerical coefficient) necessary for the conversion for C. Note that in image 3, 0 of 4 only stores the multiplier for C. Also, in image 4, the
- 3042 of 02212 is the number for the conversion in 264, ie 4X4 conversion, 4X4 conversion, 8X8 conversion, 8X8 conversion, 4X4
- a new conversion command can set a multiplier to be used in the new conversion command in accordance with the number of operations.
- the code 306 saves the new multiplier determined from the new 307 by Shina Processor.
- 0214 30 stores the numbers used by the multiplication units 3003 to 3006 corresponding to the numbers of operations. 30 is the required multiplier depending on the current value processed by the conversion table 303 and the number of operations processed by the processing unit 302, whichever is the difference between the third to the third, the third, and the third. Read from Numeral 30 sets the read multiplier to an appropriate multiplication unit 3003 to 3006.
- the multiplication units 3003 to 3006 multiply the data input to P to 4 by the number of multiplications 30 and the like.
- the calculation unit 3007 adds the data output from the multiplication unit 3003 004 and adds the data.
- the arithmetic unit 3008 adds the data output from the multiplying unit 3005 006 and adds the data.
- the output from the arithmetic unit 3007 008 is stored in the operation 3909.
- the data 3009 outputs the operation result when the calculation in the row direction and the direction is completed.
- Reference numeral 20217 denotes data for explaining the flow of the operation in the operation unit 300.
- the right hand p indicates the input.
- V, 6 on the side indicates the output.
- A, b, c, and d above indicate multipliers.
- v bX cX B and 6 dX aX B are obtained.
- the power of P is set to
- the power of 3 is set to
- the number in the multiplication unit 3003 is set to d
- the number in the multiplication is set to a
- the value output from the addition unit 3007 is dX aX R
- the force of 2 is set to
- the value output from the adder 3008 is VbXcX.
- the value output from the adder 3008 is represented by VbXcX. Therefore, it can be said that the data indicated by 20 indicates the data obtained by the calculation by the calculation unit 300.
- the arithmetic unit 300 can execute arithmetic processing capable of forming 22-power data as shown in FIG. 2 Basically, arithmetic processing that can compose two-dimensional data is used.
- the conversion of C, C4, 4X4, 8X8, 8X8, 4X4, and 2X2 can be performed. Can be performed on the conversion 3000. Therefore, the order of the operation and the operation can be defined by dividing the operation into the operation and the above-described operation processing, and by repeating the basic operation processing to the operation unit 300, the conversion can be performed. Become.
- Fig. 02212 is a diagram showing a C-converter. In 2, for example, the calculation of the part enclosed by the dotted line is explained. In 2, it is S Cos (6). Data X is input to and 2.
- S of the arithmetic unit 3004 be S.
- S of the arithmetic unit 3005 is S.
- P is output from the adder 3007.
- the calculation unit 3008 outputs P.
- the converter 3000 should input a multiplier and data to the arithmetic unit 300, and it will be explained just in case. First, one of the data toughnesses of the converter 3000 is selected out of the data toughnesses shown in FIG.
- the converter 3000 executes an operation in the row direction and the direction based on the data shown in 2 and outputs data of the conversion.
- C can be realized by turning the mark of the C-commutation data to the opposite direction, assuming that the input is on the left side and the output is on the right side without changing the multiplier. Therefore, the C-commutation may exist as a common number in 3104.
- 022522 is a diagram showing 8 ⁇ 8 conversion data. As shown in FIG. 22, even in the 8 ⁇ 8 conversion, the calculation in the row direction can be performed only by the basic data. In this way, the calculation of the direction can be performed. Therefore, 8X8 Even in this case, the multiplier required for the 8X8 conversion, which is an arithmetic function, is stored in 23.55 by the arithmetic unit 300 shown in FIG. The data 22 shown in FIG. 2 is the same except that the multiplier is different, so that the C-conversion 8X8 conversion can be performed in the same order. Therefore, the configuration of converter 3000 is simplified.
- 022623 is a figure showing 8 ⁇ 8 conversion data. As shown in FIG. 23, even in the 8X8 conversion, the calculation in the row direction can be performed only by the basic data. In this way, the calculation of the direction can be performed. Therefore, even in the 8X8 conversion, the operation is possible by the configuration of the arithmetic unit 300 shown in FIG. The multiplier required for the 8X8 conversion is stored in 2305.
- 022724 is a figure showing the data of 4X4 conversion.
- the multiplier required for the 4X4 conversion which is an arithmetic function, is stored in 2305 by the formation of the arithmetic unit 300 shown in FIG. 24 As you can see, the order of the area enclosed by the dotted line overlaps. Therefore, conversion 3000 only needs to operate so as to stop the calculation order in 8X8 conversion in the middle of 4X4 conversion.
- 022825 is a figure showing the data of the 4X4 conversion.
- the calculation in the row direction can be performed only by combining the basic data. In this way, the calculation of the direction can be performed. Therefore, even in the 4X4 conversion, the operation is possible by the configuration of the arithmetic unit 300 shown in FIG.
- the multiplier required for the 4X4 transformation is stored in 2305. 25 As you can see, the order of the area enclosed by the dotted line C is consistent, although it differs in some ways. Therefore, converter 3000 only needs to operate as a part of the input sequence in 8X8 conversion instead of 4X4 conversion.
- 022926 is a figure which shows the data of 4X4 conversion. As shown in 26
- 023027 is a figure showing the 2 ⁇ 2 conversion data. As shown in 27
- the calculation in the row direction can be calculated only by the basic data. In this way, the calculation of the direction can be performed. Therefore, even in the 2X2 conversion, the operation is possible by the configuration of the arithmetic unit 300 shown in FIG.
- the multiplier required for the 2X2 transformation is stored in 23.55.
- the design and the operation sequence for the new conversion code are divided into the above-mentioned arithmetic processing, and only the data and the numerical coefficients are set according to the number of operations. Then, as described above, create the data and determine the multiplier and the order.
- the specified multiplier is input to the new 307 according to the operation number.
- which data to read out from the data 300 and the data 3009 corresponding to the number of operations is input to the new 307.
- the new 307 converts the input multiplier to the new number 306 in accordance with the number of operations,
- the data to be read from the data 300 and the data 3009 are stored in the new memory 106.
- the new conversion method When conversion by column is performed, calculation, C and 8X8 conversion are the same, so that the operation multiplier and the destination can be easily performed.
- the multiplication unit 30 When a new calculation method is set in the command line 303, the multiplication unit 30 reads out the calculation and an appropriate multiplier from the new word line 303, makes it into the calculation unit 300, and determines which data is to be read out.
- the input selection 3002 is instructed via the section 302. In this way, even in a new conversion method, by using only the basic data to data de-tuff, it is possible to use a common method with other conversion methods.
- the arithmetic unit can be shared by reducing the conversion of the command to the combination of the basic data, so that the C, C 4X4 conversion, 4X4 conversion, and 8X8 conversion can be used. , 8X8 conversion, 4X4 conversion, 2X2 conversion, and other known conversion methods, but also miniaturized images and
- 28 is a block diagram showing the configuration of the converter 4000 according to the fourth embodiment. 28, the converter 4000 includes data 400, a plurality of calculation units 4002, an operation 4003, and operation data 4004. At 28
- the arithmetic unit 4002 has the same configuration as the arithmetic unit 300 shown in FIG. Numerical calculation units 4002 are connected in parallel.
- Reference numeral 4004 stores data obtained by the calculation by the calculation unit 4002.
- Numeral 4003 instructs the arithmetic unit 4002 on the destination and the number according to the conversion number and the number.
- the calculation unit 4003 executes the calculation of the basic table performed in the state of 3 above at a time to a plurality of calculation units 4002 so that the calculation unit 4002 outputs only the necessary multiplier and data. Instruct.
- the reference numeral 4003 indicates a control unit 30 2, a multiplier 30, a conversion register 303, and a conversion register 3043 shown in 9, a new register 30 6, and a new register 30 7.
- the arithmetic unit 4002 is instructed on the necessary multiplier and the destination of the data so that a plurality of arithmetic operations are performed on the plurality of arithmetic units 4002 at a time.
- the operation 4003 corresponds to the number of operations.
- the necessary image data is input to the difference arithmetic part of the plurality of arithmetic parts, and it is defined whether or not to perform the required operation, and the plurality of arithmetic parts are used in parallel. Since the calculation of the order can be performed for the number of numbers, the degree of the converter is improved.
- the calculation of the basic data is performed by the arithmetic unit 4002 of the number.
- the arithmetic 4003 may convert the data to be converted into the arithmetic unit 4002 of the number.
- 29 is a block diagram showing a configuration of a converter 5000 according to the fifth embodiment.
- the converter 5000 includes data 500, a plurality of arithmetic units 5002, arithmetic 5003, and arithmetic data 5004.
- data 500 stores the data.
- the arithmetic unit 5002 has the same configuration as the arithmetic unit 300 shown in FIG.
- Numerical calculation units 5002 are connected in series.
- Data 5004 stores the data obtained by the calculation by the arithmetic unit 5002 in the final stage.
- Reference numeral 5003 instructs the arithmetic unit 5002 on the destination and number of data only in accordance with the conversion data and the number.
- the arithmetic operation 5003 instructs the arithmetic units 5002 outside the first stage to use the result of the arithmetic operation of the arithmetic unit 5002.
- the operation 5003 instructs the first-stage arithmetic unit 5002 to read out the operation data 5004 or 500 data.
- the calculation unit 5002 performs the calculation of the basic position performed in step 3. While one operation unit 5002 is performing the calculation of the basic position, the other operation unit 5002 is performing another operation of the basic position. Therefore, the processing is improved because multiple operations are performed simultaneously.
- 02385003 is a logical unit 302 shown in 9, a multiplier 30, a conversion table 303, and 3 to 43045,
- the state of the state 3 is different, and the arithmetic unit 5002 is instructed on only the necessary multiplier and the destination of the data so that the plural arithmetic units 5002 are simultaneously executed.
- the operation 5003 corresponds to the number of operations.
- the necessary image data is input to the calculation part of the difference between the calculation parts, and the execution of the required calculation is defined, and the plurality of calculation parts are used in series.
- the calculation of the basic data is performed by the arithmetic unit 5002 of the number.
- the arithmetic 5003 may convert the data to be converted into the arithmetic unit 5002 of the number.
- the book may be implemented as an integrated circuit S. These books may be tipped, or some may be all tipped. Here, it is called S, but depending on the type of integration, it may also be called C, stem S, s S, or ura S.
- the integrated circuit is not limited to S, and the integrated circuit may be performed by a dedicated or general-purpose processor. It is also possible to use PG (edoabe Gaea) that can be programmed in S, or use an ngible processor that can reconfigure the settings in the S section.
- PG edoabe Gaea
- functional books may be integrated using that technology. The possibility of Io surgery is possible.
- a program capable of executing the above-described operation is stored in a computer having a device, and necessary data is stored while executing the program in P during conversion.
- the data may be stored in a device to obtain conversion data.
- the converter described in the following can be used to perform calculations specific to the 264 on the road. It is useful for articles to be worn. Further, the processing degree can be improved by performing the pipeline processing, and the present invention can be applied to uses such as (efo ee so). Also, it is possible to support C, C, commutation, inversion, and commutation, and a new conversion method.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Discrete Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Complex Calculations (AREA)
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/628,733 US8179972B2 (en) | 2004-06-08 | 2005-06-07 | Image encoding device, image decoding device, and integrated circuit used therein |
JP2006514516A JP4704333B2 (ja) | 2004-06-08 | 2005-06-07 | 画像符号化装置および画像復号化装置、ならびにそれらで用いられる集積回路 |
EP20050749005 EP1771009A1 (en) | 2004-06-08 | 2005-06-07 | Image encoding device, image decoding device, and integrated circuit used therein |
CN2005800187229A CN1981534B (zh) | 2004-06-08 | 2005-06-07 | 图像编码装置、图像解码装置以及用于其中的集成电路 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2004169726 | 2004-06-08 | ||
JP2004-169726 | 2004-06-08 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2005122590A1 true WO2005122590A1 (ja) | 2005-12-22 |
Family
ID=35503517
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2005/010419 WO2005122590A1 (ja) | 2004-06-08 | 2005-06-07 | 画像符号化装置および画像復号化装置、ならびにそれらで用いられる集積回路 |
Country Status (5)
Country | Link |
---|---|
US (1) | US8179972B2 (ja) |
EP (1) | EP1771009A1 (ja) |
JP (1) | JP4704333B2 (ja) |
CN (1) | CN1981534B (ja) |
WO (1) | WO2005122590A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009159287A (ja) * | 2007-12-26 | 2009-07-16 | Fujitsu Ltd | 半導体装置および信号処理方法 |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100771879B1 (ko) * | 2006-08-17 | 2007-11-01 | 삼성전자주식회사 | 내부 메모리 용량을 감소시키는 디블록킹 필터링 방법 및그 방법을 이용하는 영상 처리 장치 |
US20080288568A1 (en) * | 2007-05-14 | 2008-11-20 | Hou Hsieh S | Low power Fast Hadamard transform |
CN102656884A (zh) * | 2009-12-16 | 2012-09-05 | 国际商业机器公司 | 使用像素流执行视频编码 |
CN102314528B (zh) * | 2010-07-07 | 2013-05-01 | 中国科学院微电子研究所 | 一种数字大规模集成电路的资源共享系统及方法 |
US9055304B2 (en) * | 2011-07-01 | 2015-06-09 | Qualcomm Incorporated | Reduced resolution pixel interpolation |
CN106204680A (zh) * | 2016-07-18 | 2016-12-07 | 京东方科技集团股份有限公司 | 一种成像装置及其成像方法 |
CN108848388B (zh) * | 2018-07-17 | 2022-02-22 | 珠海亿智电子科技有限公司 | 一种提高H264编码16x16预测模式DCT运算速度的硬件实现方法 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05181896A (ja) * | 1991-12-27 | 1993-07-23 | Nec Corp | 離散コサイン変換処理装置 |
JPH0646404A (ja) * | 1992-03-17 | 1994-02-18 | Sony Corp | 画像データ処理装置及び画像データ処理方法 |
US5471412A (en) * | 1993-10-27 | 1995-11-28 | Winbond Electronic Corp. | Recycling and parallel processing method and apparatus for performing discrete cosine transform and its inverse |
JPH0844708A (ja) * | 1994-07-27 | 1996-02-16 | Fujitsu Ltd | 二次元離散コサイン変換演算回路 |
JP2000099495A (ja) * | 1998-09-15 | 2000-04-07 | Winbond Electron Corp | 乗法演算節減可能な二次元離散コサイン変換及びその逆変換の実行方法 |
JP2001325247A (ja) * | 2000-05-17 | 2001-11-22 | Nippon Telegr & Teleph Corp <Ntt> | 高速逆離散コサイン変換装置及び高速逆離散コサイン変換方法及び当該方法を記述したプログラムを記録した記録媒体 |
JP2004120439A (ja) * | 2002-09-26 | 2004-04-15 | Nec Electronics Corp | 画像処理装置及び画像処理方法 |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5168375A (en) * | 1991-09-18 | 1992-12-01 | Polaroid Corporation | Image reconstruction by use of discrete cosine and related transforms |
EP0561593B1 (en) * | 1992-03-17 | 1997-07-16 | Sony Corporation | Image compression apparatus |
JP3069455B2 (ja) * | 1992-12-22 | 2000-07-24 | 富士写真フイルム株式会社 | 画像データ圧縮伸張装置における量子化・逆量子化回路 |
JPH0723381A (ja) | 1993-06-23 | 1995-01-24 | Nec Corp | 画像の復号化方法及びその復号化装置 |
US6157740A (en) * | 1997-11-17 | 2000-12-05 | International Business Machines Corporation | Compression/decompression engine for enhanced memory storage in MPEG decoder |
US6587590B1 (en) * | 1998-02-02 | 2003-07-01 | The Trustees Of The University Of Pennsylvania | Method and system for computing 8×8 DCT/IDCT and a VLSI implementation |
CN1147155C (zh) * | 1998-12-14 | 2004-04-21 | 松下电器产业株式会社 | Dct运算装置 |
JP4551997B2 (ja) | 1999-10-25 | 2010-09-29 | 独立行政法人情報通信研究機構 | 画像の符号化および復号方法と装置 |
JP3643777B2 (ja) * | 2001-02-28 | 2005-04-27 | 三洋電機株式会社 | 動画像復号装置および動画像復号方法 |
US7123655B2 (en) | 2001-08-09 | 2006-10-17 | Sharp Laboratories Of America, Inc. | Method for reduced bit-depth quantization |
US7295609B2 (en) * | 2001-11-30 | 2007-11-13 | Sony Corporation | Method and apparatus for coding image information, method and apparatus for decoding image information, method and apparatus for coding and decoding image information, and system of coding and transmitting image information |
JP4348899B2 (ja) | 2001-11-30 | 2009-10-21 | ソニー株式会社 | 画像情報復号方法 |
JP2003223433A (ja) * | 2002-01-31 | 2003-08-08 | Matsushita Electric Ind Co Ltd | 直交変換方法、直交変換装置、符号化方法、符号化装置、逆直交変換方法、逆直交変換装置、復号化方法、及び、復号化装置 |
JP3902990B2 (ja) * | 2002-07-02 | 2007-04-11 | キヤノン株式会社 | アダマール変換処理方法及びその装置 |
-
2005
- 2005-06-07 WO PCT/JP2005/010419 patent/WO2005122590A1/ja active Application Filing
- 2005-06-07 JP JP2006514516A patent/JP4704333B2/ja not_active Expired - Fee Related
- 2005-06-07 US US11/628,733 patent/US8179972B2/en not_active Expired - Fee Related
- 2005-06-07 CN CN2005800187229A patent/CN1981534B/zh not_active Expired - Fee Related
- 2005-06-07 EP EP20050749005 patent/EP1771009A1/en not_active Withdrawn
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05181896A (ja) * | 1991-12-27 | 1993-07-23 | Nec Corp | 離散コサイン変換処理装置 |
JPH0646404A (ja) * | 1992-03-17 | 1994-02-18 | Sony Corp | 画像データ処理装置及び画像データ処理方法 |
US5471412A (en) * | 1993-10-27 | 1995-11-28 | Winbond Electronic Corp. | Recycling and parallel processing method and apparatus for performing discrete cosine transform and its inverse |
JPH0844708A (ja) * | 1994-07-27 | 1996-02-16 | Fujitsu Ltd | 二次元離散コサイン変換演算回路 |
JP2000099495A (ja) * | 1998-09-15 | 2000-04-07 | Winbond Electron Corp | 乗法演算節減可能な二次元離散コサイン変換及びその逆変換の実行方法 |
JP2001325247A (ja) * | 2000-05-17 | 2001-11-22 | Nippon Telegr & Teleph Corp <Ntt> | 高速逆離散コサイン変換装置及び高速逆離散コサイン変換方法及び当該方法を記述したプログラムを記録した記録媒体 |
JP2004120439A (ja) * | 2002-09-26 | 2004-04-15 | Nec Electronics Corp | 画像処理装置及び画像処理方法 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009159287A (ja) * | 2007-12-26 | 2009-07-16 | Fujitsu Ltd | 半導体装置および信号処理方法 |
JP4536109B2 (ja) * | 2007-12-26 | 2010-09-01 | 富士通株式会社 | 半導体装置および信号処理方法 |
Also Published As
Publication number | Publication date |
---|---|
EP1771009A1 (en) | 2007-04-04 |
JPWO2005122590A1 (ja) | 2008-04-10 |
CN1981534B (zh) | 2011-04-13 |
CN1981534A (zh) | 2007-06-13 |
JP4704333B2 (ja) | 2011-06-15 |
US8179972B2 (en) | 2012-05-15 |
US20080049832A1 (en) | 2008-02-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP2672302B2 (ja) | 変換回路 | |
WO2005122590A1 (ja) | 画像符号化装置および画像復号化装置、ならびにそれらで用いられる集積回路 | |
JP4874642B2 (ja) | ロッシーおよびロスレス2dデータ圧縮のための可逆な変換 | |
JPH09198373A (ja) | 逆離散コサイン変換プロセッサおよびmpegデコーダ | |
JP3796432B2 (ja) | フィルタ処理装置およびフィルタ処理方法 | |
JPH0799578A (ja) | 画像処理装置 | |
JPH09198371A (ja) | パラレルデータワード・シリアルワード変換装置および分配演算装置 | |
JPH0645948A (ja) | 直交変換装置及び逆直交変換装置 | |
JPWO2005099249A1 (ja) | 圧縮装置、色変換装置、その方法、プログラム、ルックアップテーブル及び記憶媒体 | |
JP4266512B2 (ja) | データ処理装置 | |
JPH0645947A (ja) | 直交変換装置及び逆直交変換装置 | |
JP3046115B2 (ja) | 離散コサイン変換器 | |
JP3385866B2 (ja) | 逆量子化・逆dct回路 | |
JP3046116B2 (ja) | 離散コサイン変換器 | |
JPH0646269A (ja) | 静止画像データの伸長方法、圧縮方法及びそのための対応装置 | |
JP3366033B2 (ja) | アドレス変換回路及びこれを備えた直交変換回路 | |
JP5451171B2 (ja) | データ変換処理装置およびデータ変換処理方法 | |
JPH04531A (ja) | 直交変換演算装置 | |
JPH0693753B2 (ja) | 画像変換符号化装置 | |
JPH09198372A (ja) | キャリー値発生装置および加減算mビット値発生装置 | |
JPH04211575A (ja) | 直交変換演算装置 | |
JP2923527B2 (ja) | 画像データ符号化・復元装置 | |
JPH03270376A (ja) | 画像データ直交変換処理装置のマトリクス演算回路 | |
JPH01198882A (ja) | テレビジョン信号の高能率符号化方式のデコード装置 | |
JPH1049519A (ja) | 離散コサイン変換器 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KM KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NG NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SM SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2006514516 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 11628733 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 200580018722.9 Country of ref document: CN |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWW | Wipo information: withdrawn in national office |
Country of ref document: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2005749005 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 2005749005 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 11628733 Country of ref document: US |