WO2016129031A1 - 画像圧縮伸長装置 - Google Patents
画像圧縮伸長装置 Download PDFInfo
- Publication number
- WO2016129031A1 WO2016129031A1 PCT/JP2015/053501 JP2015053501W WO2016129031A1 WO 2016129031 A1 WO2016129031 A1 WO 2016129031A1 JP 2015053501 W JP2015053501 W JP 2015053501W WO 2016129031 A1 WO2016129031 A1 WO 2016129031A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- processing unit
- variable length
- parallel
- length coding
- Prior art date
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
- H04N19/42—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
- H04N19/436—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 using parallelised computational arrangements
-
- 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
- H04N19/48—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data
-
- 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
- H04N19/127—Prioritisation of hardware or computational resources
-
- 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
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/136—Incoming video signal characteristics or properties
-
- 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
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/146—Data rate or code amount at the encoder output
-
- 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
-
- 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
- H04N19/40—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using video transcoding, i.e. partial or full decoding of a coded input stream followed by re-encoding of the decoded output stream
-
- 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
- H04N19/44—Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder
Definitions
- the present invention relates to an image compression / decompression apparatus.
- the present invention relates to an image compression / decompression apparatus suitable for operating coding / decoding processing of image processing based on video compression coding technology such as H.264 standard with low power consumption.
- a technique for compressing (encoding) and decompressing (decoding) an image using an MPEG compression technique or the like is used in digital image processing equipment.
- ISO / IEC standard 23008-3 commonly known as H.265 / HEVC
- H.265 / HEVC High Efficiency Video Coding
- the image encoding device described in Patent Document 1 is provided with a buffer between the binarization unit 301 and the binary arithmetic encoding unit 304, and the processing clocks of the binarization unit 301 and the binary arithmetic encoding unit 304 are different.
- the power consumption can be reduced by assigning an appropriate operating frequency to each processing unit.
- An object of the present invention is to provide an image compression / decompression apparatus that can reduce power consumption in accordance with a user's usage, considering resolution, frame rate, and bit rate.
- the image compression / decompression apparatus is preferably an apparatus that performs image compression / decompression processing for compressing original image data based on encoding, and decompressing the encoded data.
- the first parallel number is determined based on the resolution or frame rate of the image, and the one or more image processing units are configured to perform parallel processing with a parallel degree corresponding to the first parallel number.
- the second parallel number is determined based on the bit rate of the bit stream processed by the image compression / decompression apparatus, and the one or more variable length coding / decoding processing units are set to the second parallel number. Parallel processing is performed according to the degree of parallelism.
- the one or more image processing units are input with both the image resolution and the frame rate, or one of the image resolution and the frame rate as a parameter, in order to determine the first parallel number.
- an image compression / decompression apparatus capable of reducing power consumption in accordance with a user's usage application in consideration of resolution, frame rate, and bit rate.
- 1 is an overall configuration diagram of a configuration of an image compression / decompression apparatus according to an embodiment. It is a figure explaining the mode which determines the parallel number of an image process part and a variable-length encoding decoding process part. It is a figure explaining the relationship between an image (1 frame) when a parallel number of an image processing part is 1, and a macroblock. It is a figure which shows a mode that an image process part processes a macroblock for every processing stage by a time unit. It is a figure which shows a mode when data is transferred from an external memory to an image processing part and a variable-length encoding decoding process part at the time of 1 image processing part system at the time of compression operation, and 1 variable-length encoding decoding process part operation
- FIG. 1 is an overall configuration diagram of an image compression / decompression apparatus.
- the image compression / decompression device is H.264.
- H.264 (ISO / IEC14496-10) is a circuit that performs image compression / decompression processing, and is a decompression device that performs half-duplex processing of image compression processing and image decompression processing.
- an original image (YCbCr 4: 2: 0 format) is input and sequentially stored in an original image buffer (not shown in FIG. 1) arranged in the external memory 105, and the original image is encoded. It is a process to compress.
- the image decompression process is performed by inputting a bit stream sequentially stored from MSB (Most Significant Bit) to a stream buffer (not shown in FIG. 1) arranged in the external memory 105, and decoding the image.
- a decoded image (YCbCr 4: 2: 0 format) is output and sequentially stored in a decoded image buffer arranged in the external memory 105.
- the configuration of the present apparatus is divided into an image processing unit that is responsible for motion compensation processing when performing image compression / decompression processing, and a modulation and coding decoding processing unit that performs entropy coding processing.
- the image processing units 101, 102, 103, 104 and the variable length coding / decoding processing units 124, 125 are connected via an intermediate buffer area (not shown in FIG. 1) provided in the external memory 105. Deliver data.
- the image processing unit 101 is activated from the image processing unit controller 122.
- the image processing unit 101 stores the original image in the original image buffer unit 106 from the external memory 105 via the bus 121.
- the intra prediction mode determination unit 108 performs prediction mode determination, and generates an intra prediction image.
- the search unit 107 calculates a motion vector optimal for inter prediction, and a reference image at a position corresponding to the motion vector is obtained from a reference frame area (not shown in FIG. 1) in the external memory 105 by the bus 121.
- the residual calculation part 110 produces
- the prediction error is input to the next discrete cosine transform unit 111.
- the input residual component is subjected to integer cosine transform, converted into frequency samples, and output to the quantization block unit 112.
- the input data is divided by a predetermined quantization scale.
- Each frequency sample after quantization is read out from the two-dimensional matrix by the zigzag scanning unit 113 in the zigzag scanning order, and is output as one-dimensional data.
- the residual calculation unit 110 transfers the macroblock division information and the prediction mode in the case of intra prediction, and the motion vector to the run length encoding unit 114 in the case of inter prediction. Thereafter, the run-length encoding unit 114 performs H.264 encoding.
- the data is converted into data having a continuous number of zeros and non-zero coefficients and is provided in an intermediate buffer area (not shown in FIG.
- each frequency sample divided by the quantization unit 112 to generate a reference image is multiplied by a predetermined quantization scale by the inverse quantization unit 117, and the frequency sample is inversely quantized. It becomes.
- an inverse discrete cosine transform unit 118 performs integer cosine transform on the frequency sample and transforms it into a residual component.
- an intra prediction image is generated from the prediction mode by the intra prediction unit 109, and in the case of inter prediction, a reference frame at a position corresponding to the motion vector is stored in the external memory 105 by the search unit 107.
- An inter prediction image is generated by reading out from the region (not shown in FIG. 1) via the bus 121.
- the residual calculation unit 119 adds the intra prediction image or the inter prediction image and the residual component to generate a prediction image. Thereafter, the deblocking filter 120 removes block distortion to generate a reference image, which is stored in the external memory 105 via the bus 121.
- image decompression processing (hereinafter, simply referred to as “decompression processing”) of the image compression / decompression apparatus will be described.
- variable length coding / decoding processor controller 130 activates the variable length coding / decoding processor 124.
- the variable length coding / decoding processing unit 124 reads a bit stream from a stream buffer (not shown in FIG. 1) arranged in the external memory 105 via the bus 121 and stores the bit stream in the code buffer unit 127.
- the variable length decoding unit 128 performs H.264. In accordance with the H.264 standard, the number of consecutive 0s and non-zero coefficient data, macroblock division information, intra prediction mode, or motion vector are decoded.
- the decoded data is an intermediate buffer (see FIG. 4) that is formed in a form that can be understood when the run (appearance of symbols) and length (length of symbol arrangement) are read later according to a predetermined format. 1 (not shown in FIG. 1) via the bus 121.
- variable length coding / decoding processing unit controller 130 After the processing for one frame is completed in the variable length coding / decoding processing unit 124, the variable length coding / decoding processing unit controller 130 notifies the image processing unit controller 122.
- the image processing unit controller 122 activates the image processing unit 101, and the run length decoding unit 115 of the image processing unit 101 acquires intermediate data from the external memory 105 via the bus 121.
- the run-length decoding unit 115 calculates all frequency samples in the macroblock from the continuous number of zeros and the non-zero coefficient data, and sends it to the zigzag scanning unit 116.
- the zigzag scan unit 116 outputs the one-dimensional data as a two-dimensional matrix in the zigzag scan order.
- the inverse quantization unit 117 performs multiplication by a predetermined quantization scale to inversely quantize the frequency sample.
- an inverse discrete cosine transform unit 118 performs integer cosine transform on the frequency sample and transforms it into a residual component.
- the intra prediction unit 109 generates an intra prediction image according to the prediction mode.
- the search unit 107 refers to a reference image in the external memory 105 at a position corresponding to the motion vector. Reading from the frame region (not shown in FIG. 1) via the bus 121 generates an inter prediction image.
- the intra prediction image or the inter prediction image is added by using the residual component in the residual calculation unit 119 to generate a prediction image.
- the block distortion is removed by the deblocking filter 120 to generate a decoded image, which is stored in the decoded image buffer (not shown in FIG. 1) disposed in the external memory 105 via the bus 121.
- the decoded image in the decompression process is also used as a reference image for the decompression process.
- the image compression / decompression apparatus has a plurality of image processing units and a variable length coding / decoding processing unit, and is premised on that they operate independently and in parallel. Therefore, the parallel number of the image processing units 101, 102, 103, and 104 and the variable length coding / decoding processing units 124 and 125 will be described with reference to FIG.
- FIG. 2 is a diagram for explaining a mode for determining the parallel number of the image processing units 101, 102, 103, 104 and the variable length coding / decoding processing units 124, 125.
- the processing time of the image processing unit is almost specified depending on the pixel rate calculated by resolution ⁇ frame rate.
- the resolution is a numerical value indicating the density of pixels in the bitmap image
- the frame rate is the number of frames (the number of still images and the number of frames) processed per unit time in the moving image.
- the image processing unit 101 for one system shown in FIG. 1 is capable of processing a resolution of 1920 ⁇ 1088 and a frame rate of 30 fps
- the resolution is 1920 ⁇ 1088 and the frame rate is 60 fps
- processing with a resolution of 4096 ⁇ 2160 and a frame rate of 30 fps can be performed.
- the frame rate may be 120 fps when the resolution is 1920 ⁇ 1088.
- the image processing unit determines the parallel number based on the pixel rate calculated by resolution ⁇ frame rate, and the image processing units 101, 102, 103, and 104 according to the degree of parallelism according to the parallel number. Are operated in parallel.
- the processing time of the processing of the variable length encoding / decoding processing unit varies depending on the bit rate of the bit stream.
- the bit rate is a numerical value indicating how many bits of data are processed or transmitted / received per unit time.
- the variable length coding / decoding processing unit 124 is assumed to have a processing performance of 50 Mbps, and the parallel number is defined by the bit rate transferred to the variable length coding / decoding processing unit 124. Then, the variable length coding / decoding processing units 124 and 125 cause the variable length coding / decoding processing units 124 and 125 to operate in parallel according to the degree of parallelism according to the number of parallels defined by the bit rate. Thereby, it becomes possible to perform easy electric power control according to a use application.
- resolution 4096 ⁇ 2160
- frame rate 30 fps
- bit rate 50 Mbs to 100 Mbs
- the parallel number of image processing units 4
- the image processor controller 122 has a table for determining the number of parallel image processors when the resolution and frame rate are set in advance, as shown in FIG. .
- mode no. 2 the image processing unit is operated with a parallel number of parallel number 2
- the mode No. 5 the image processing unit is operated at a parallel degree of 4 parallels.
- variable length coding / decoding processing unit controller 130 is connected in parallel to the variable length coding / decoding processing unit.
- the image processing unit controller 122 and the variable length coding / decoding processing controller 130 prevent unnecessary data paths from changing for each clock according to the operation mode requested by the user, and consumption during operation of the image compression / decompression apparatus. Electric power can be reduced.
- the resolution / frame rate and the bit rate may both be input variably, or the resolution / frame rate may be changed, the parameters may be input, the bit rate may be fixed, and vice versa. Furthermore, the resolution and frame rate may be fixed, and the bit rate may be input as a parameter.
- the parallel number of the image processing units is determined according to the resolution and the frame rate, and the parallel number of the variable length coding / decoding processing unit is determined according to the bit rate.
- the same low power effect can be obtained even if the register setting is made directly from an external host CPU or the like.
- FIG. 3 is a diagram for explaining the relationship between an image (one frame) and a macroblock when the number of parallel image processing units is one.
- the image processing unit 101 sequentially performs the right horizontal direction (MB0, MB1, MB2,%) On the input image from the upper left macroblock MB0 for each unit called a 16 ⁇ 16 pixel macroblock. 7) Reads the original image and performs macroblock compression processing.
- the macroblock line (MB119) for one horizontal line ends from MB0 the image processing unit 100 moves in the vertical direction of the lower 16 pixels and sequentially moves from the left end (MB120) to the right in the horizontal direction. The process is repeated until the frame is completed.
- the original image is sequentially transferred from the image input unit (not shown) to the original image buffer (described later in FIG. 5) of the external memory 105 in real time as (B0, B1, I2, B3, B4, P5,). Shall be entered.
- “I” is an I picture (intra-frame encoded image)
- “P” is a P picture (inter-frame forward predictive encoded image)
- “B” is a B picture (bidirectional predictive coded image), and the number promises to represent the input frame order when counting from the 0th frame of the original image.
- the above description describes a series of flow of the macroblock processing, and does not include sub-information such as adjacent macroblock information.
- FIG. 4 is a diagram illustrating a state in which the image processing unit 101 processes a macroblock in units of time for each processing stage.
- the original image reading stage is the original image buffer unit 106
- the intra prediction mode determination stage is the intra prediction mode determination unit 108
- the intra inter determination stage is the intra prediction unit 109
- the search unit 107 and the prediction image creation stage.
- the residual calculation unit 110 the frequency transformation stage is the discrete cosine transformation unit 111, the quantization 112, the inverse quantization 117, the inverse discrete cosine transformation unit 118
- the run length stage is the zigzag scanning unit 113, and the run length.
- the encoding unit 114, the decoded image creation stage are the residual calculation unit 119
- the deblocking filter stage is a stage at which the deblocking filter 120 operates.
- Time 0 (time 1, time 2,%) Indicates a processing time required for each MB.
- time 0 (time 1, time, time)
- Each stage ends within 2...
- the macro block sequentially reads the original image from the upper left of the image (MB0, MB1, MB2,). Thereafter, the processing stage is pipeline-operated such as intra prediction mode determination and intra-inter determination. For example, at time 3, a macroblock of MB3 for the original image read stage, MB2 for the intra prediction mode determination stage, MB1 for the intra inter determination stage, and MB0 for the prediction image creation stage is processed.
- the image processing unit 101 repeats the macroblock processing until the processing for one frame is completed as shown in FIG.
- the image processing unit controller 122 notifies the variable length coding / decoding processing controller 130 after completion of one frame.
- the variable length coding / decoding processing unit controller 130 activates the variable length coding / decoding processing unit 124.
- the variable length encoding / decoding processing unit 124 reads the intermediate data stored in the image processing unit 101 from the external memory 105 unit via the bus 121, generates a bit stream compressed by the variable length encoding unit 126, and generates a code buffer.
- the unit 127 sequentially stores bits from the MSB in the stream buffer arranged in the external memory 105 via the bus 121. Thus, in the compression process, the process is performed for each frame until the process is completed, and the compressed bit stream is sequentially stored in the stream buffer.
- FIG. 5 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit when one image processing unit and variable length coding / decoding processing unit are operated during the compression operation.
- the image processing unit 101 inputs images from the original image buffer 300 in the compression order (I2, B0, B1,...), And sequentially stores intermediate data in the intermediate buffer 310.
- the variable length encoding processing unit 124 reads and compresses the intermediate data from the intermediate buffer 310 and sequentially stores the generated bit stream in the stream buffer 320.
- FIG. 6 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit during one operation of the image processing unit and variable length coding / decoding processing unit during the decompression operation.
- FIG. 6 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit during one operation of the image processing unit and variable length coding / decoding processing unit during the decompression operation.
- the variable length coding processing unit 124 acquires a bit stream from the stream buffer 330 and sequentially stores intermediate data in the intermediate buffer 310.
- the image processing unit 101 reads out the intermediate data from the intermediate buffer 310, performs decompression processing, and sequentially stores the generated decoded images in the decoded image buffer 340.
- FIG. 7 is a diagram for explaining the relationship between an image (one frame) and a macroblock when the number of parallel image processing units is four.
- FIG. 8 is a diagram illustrating a state in which the image processing units 101, 102, 103, and 104 process macroblocks in units of time for each processing stage.
- pipeline processing is performed in the same manner as in one system for each of the divided different macroblocks.
- the macroblock reads the original image sequentially from the MB0, the image processing unit 102 from the MB8704, the image processing unit 103 from the MB17408, and the image processing unit 104 from the MB2612. Thereafter, a pipeline operation is performed for each processing stage, such as intra prediction mode determination and intra-inter determination.
- the original image reading stage is MB3, MB8707, MB17411, MB26115
- the intra prediction mode determination stage is MB2, MB8706, MB17410, MB26114, intra for each of the image processing units 101, 102, 103, and 104.
- the inter determination stage processes MB1, MB8705, MB17409, MB26113
- the prediction image creation stage processes macroblocks MB0, MB8704, MB17408, and MB26112.
- FIG. 9 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit when the four image processing units and the variable length coding / decoding processing unit are operated in the compression operation.
- FIG. 9 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit when the four image processing units and the variable length coding / decoding processing unit are operated in the compression operation.
- the image processing units 101, 102, 103, and 104 input images from the image buffer 300 in the compression order (I2, B0, B1,...), And sequentially store the intermediate data in the intermediate buffer 310 that is divided into four.
- the image processing unit 101 is 710 of the intermediate buffer 310
- the image processing unit 102 is 711 of the intermediate buffer 310
- the image processing unit 103 is 712 of the intermediate buffer 310
- the image processing unit 104 is Intermediate data is stored in 713 of the intermediate buffer 310, respectively.
- the image processing unit 101 is 714 of the intermediate buffer 310
- the image processing unit 102 is 715 of the intermediate buffer 310
- the image processing unit 103 is 716 of the intermediate buffer 310
- the image processing unit 104 is Intermediate data is stored in 717 of the intermediate buffer 310, respectively.
- the data is sequentially stored in the intermediate buffer 310 divided into four, such as “B1 (P5, B3, B4...)”.
- variable length encoding processing units 124 and 125 read intermediate data from the intermediate buffer 310 in the order of compression.
- the variable length encoding / decoding processing unit 124 stores the compressed bit stream in the intermediate buffer 320 and the variable length encoding / decoding processing unit 125. Are sequentially stored in the intermediate buffer 321 in units of frames. For example, as shown in FIG.
- variable length coding processing unit 124 processes the original images 702 and 703
- variable length coding processing unit 125 processes the original images 700, 701, 704, and 705, respectively
- the bit stream 722 of the original image 702 and the bit stream 723 of the original image 703 are stored in the buffer 320
- the bit stream 720 of the original image 700, the bit stream 721 of the original image 701, and the bit stream of the original image 704 are stored in the intermediate buffer 321.
- the bit stream 725 of the original image 705 is stored.
- the intermediate buffers 320 and 321 are mixed in units of frames. This alignment can be solved by reading from the intermediate buffers 320 and 321 in the compression order in the stream combination analysis module 129 described later, combining them, and storing the bit stream in the stream buffer 330.
- FIG. 10 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit during the operation of the four image processing units and the variable length coding / decoding processing unit during the decompression operation.
- FIG. 10 shows a state in which data is transferred from an external memory to the image processing unit and the variable length coding / decoding processing unit during the operation of the four image processing units and the variable length coding / decoding processing unit during the decompression operation.
- the stream combination analysis module 129 performs an analysis for cutting out the compressed data in the stream buffer 330 in units of one frame. Based on the analysis result, the variable length encoding processing units 124 and 125 read from the stream buffer 330 in units of frames, and the variable length encoding and decoding processing unit 124 stores the intermediate data in the intermediate buffer 310 and the variable length encoding and decoding processing unit. 125 are sequentially stored in the intermediate buffer 311.
- the image processing units 101, 102, 103, and 104 select the intermediate buffers 310 and 311 in the frame order and read the intermediate data.
- the generated decoded images are stored in the decoded image buffer 340 in the display order (B0, B1, I2,).
- the image processing unit and the variable length coding / decoding unit can know how much data is stored by notifying the buffer usage amount and can be used as a ring buffer. In addition, by performing such buffer control, it is possible to buffer data between an image processing unit and a variable-length encoding / decoding unit having different processing units, and to realize real-time compression processing or decompression processing. It becomes.
- FIG. 11 is a flowchart showing a detailed procedure of image compression processing of the image compression / decompression apparatus.
- FIG. 12 is a timing chart showing the processing timing of each picture and each part in the image compression processing.
- the image processing unit controller 122 performs initial setting for the image processing units 101,... According to the compression start instruction from the start instruction waiting loop (S01) (S02). In this initial setting, in which mode the image processing units 101,... Operate, that is, the parallel number of the image processing units 101,. Then, the image processing unit controller 122 issues a variable length coding / decoding start instruction to the variable length coding / decoding processing unit controller 130 (S03, A1), and the initial setting from the variable length coding / decoding processing unit controller 130 is completed. Wait for notification (S04).
- variable length coding / decoding processing unit controller 130 shifts to the initial setting from the start instruction waiting loop (S11) (S12), and upon completion, notifies the image processing unit controller 122 of the initial setting completion (S13, A2).
- this initial setting in which mode the variable-length coding / decoding processing units 124,... Operate, that is, the parallel number of the variable-length coding / decoding processing units 124,.
- the image processing unit 101 Repeats the compression process in units of one frame until an end instruction is issued (S05, S07), and notifies the variable length coding processing unit controller 130 of one frame completion each time the process is completed ( S06, A3).
- the variable length coding / decoding processing unit 124 receives the completion notification from the image processing unit 101,... Starts compression processing for one frame (S14), and repeats until there is an end notification (S15).
- the variable-length coding processing unit controller 130 may receive a one-frame completion notification from the image processing unit controller 122 while the variable-length coding processing unit is performing one-line operation processing. In this case, the notification may be retained and the retained frame may be processed after completion.
- variable-length coding processing unit when the variable-length coding processing unit is operating in two systems, variable-length coding that is not being processed even if one frame is being processed, as shown in the timing chart of the compression processing shown in FIG.
- the decoding processing unit By causing the decoding processing unit to process, two frames can be processed simultaneously.
- FIG. 12 if the variable length coding / decoding processing unit 1 is processing “I2”, the variable length coding / decoding processing unit 2 processes “B0”, “B1”, and “P5”. Thereafter, the stream combination analysis module 129 described later connects in the order of compression, so that parallel operations can be performed.
- variable length coding / decoding processing unit may receive a one frame completion notification from the image processing unit during the two-line operation processing. In this case, holding is performed in the same manner as the one-line operation, and the held frame is processed after completion. In the two-system operation, the order of completion depends on the bit amount of each frame. In this embodiment, the function of concatenating the stream data in the frame order is provided, so that the frame order (I2, B0, B1, P5, B3, B4,...) Can be generated.
- FIG. 13 is a flowchart showing a detailed procedure of image decompression processing of the image compression / decompression apparatus.
- FIG. 14 is a timing chart showing the processing timing of each picture and each part in the image expansion processing.
- the image processing unit controller 122 makes initial settings for the image processing units 101,... In response to a decompression start instruction from the start instruction waiting loop (S21) (S22). In this initial setting, in which mode the image processing units 101,... Operate, that is, the parallel number of the image processing units 101,. Then, the image processing unit controller 122 issues a variable length coding / decoding start instruction to the variable length coding / decoding processing unit controller 130 (S23, A11), and the initial setting from the variable length coding / decoding processing unit controller 130 is completed. Wait for notification (S24).
- variable length coding / decoding processing unit controller 130 shifts to the initial setting from the start instruction waiting loop (S31) (S32), and upon completion, notifies the image processing unit controller 122 of the initial setting completion (S33, A12).
- this initial setting in which mode the variable-length coding / decoding processing units 124,... Operate, that is, the parallel number of the variable-length coding / decoding processing units 124,.
- variable length coding / decoding processing unit 124 Repeats the decompression processing in units of one frame until an end notification is received (S34, S36), and notifies the image processing unit of one frame completion every time the processing is completed (S35, S36). A13).
- the image processing unit 101 Receives a completion notification from the variable length coding / decoding processing unit, starts an expansion process for one frame, and repeats until an end instruction is given (S25, S26, S27).
- variable-length coding / decoding processor controller In the 2-system operation, the variable-length coding / decoding processor controller must allocate variable-length coding / decoding processing in units of one frame. As shown in the timing chart of the decompression process in FIG. 14, the stream data is analyzed, and the frame unit process can be performed even in a two-line operation by detecting the head of the frame. If the variable length coding / decoding processing unit 1 is processing “I2”, the variable length coding / decoding processing unit 2 processes “B0”, “B1”, and “P5”.
- the completion order of each frame may be changed depending on the amount of bits to be expanded. Since the variable-length encoding / decoding processor controller 130 allocates and manages processes for each system in the order of frames, in such a case, the variable-length encoding / decoding processor controller 130 performs image processing in the order of frames. By performing the notification, the decompression process can be performed without changing the frame order (I2, B0, B1, P5, B3, B4,%) Of the image processing unit.
- the stream combination analysis block unit 129 is a part that functions when, for example, the variable-length encoding / decoding processing units 124 and 125 operate in parallel. In the compression processing, the variable-length encoding / decoding processing units 124 and 125 respectively generate Combine stream data in units of one frame.
- the stream combination analysis block unit 129 notifies the host CPU (not shown) of frame breaks in advance, and sequentially reads the stream of each frame on the host side and combines them on software. It is good also as processing. Further, at the time of decompression processing, analysis is performed to cut out compressed stream data to be decompressed by one frame which is a processing unit of the variable length coding / decoding processing units 124 and 125.
- an AU delimiter header is added to the beginning of each frame, and if a unique AU delimiter code (0x09) is added to the beginning of each picture, this unique ID is searched during analysis. It becomes possible to determine the frame break. Further, in this analysis process, a frame break is detected in advance by a host CPU (not shown), each frame break information is notified to the circuit, and the variable length coding / decoding processing units 124 and 125 are detected. However, each may be read from the external stream buffer based on the frame break information.
- FIG. 15A is a diagram illustrating an example of controlling parallel processing of the image processing unit and the variable length coding / decoding processing unit by clock control.
- FIG. 15B is a diagram illustrating an example of controlling parallel processing of the image processing unit and the variable length coding / decoding processing unit by power control.
- the external clock is gated by the image processing units 101, 102, 103, 104 and the variable length coding / decoding processing units 124, 125 in the clock control unit of the image compression / decompression apparatus. And distribute to each processing unit.
- the clock control units 123 and 131 each have a CG (clock gating cell) connected to the image processing unit controller 122 and the variable length coding / decoding processing controller 130, respectively.
- CG clock gating cell
- the clock tree (component connected to the clock) distributed to each processing unit after clock gating has a large fanout (the maximum number of circuit elements that can be connected)
- only the clock is supplied.
- a large amount of power is consumed.
- a large power consumption is achieved by controlling the clocks from the clock control units 123 and 131 to the unused blocks according to the operation mode selection of the image processing unit controller 122 and the variable length coding / decoding processing controller 130, respectively. A reduction effect can be produced.
- the power for stopping the power supply of the image processing unit and the variable length coding / decoding processing unit that are not operating among the gates in the circuit is provided, and the power control unit is switched according to the presence or absence of operation.
- a logic gate may be inserted so that a definite value is input.
- the image processing unit corresponds to the pixel rate calculated by the resolution at the time of operation ⁇ the frame rate.
- CABAC Context-Adaptive-Binary-Arithmetic-Coding
- CAVLC Context-Adaptive-Variable-Length-Coding
- CABAC is a coding method that performs a process of changing the prediction model of the probability of occurrence of the next bit according to the occurrence of each bit of binarized data. For this reason, sequential processing is performed on a continuous data string in one slice, and the processing time greatly varies depending on the number of bits to be processed. Therefore, even if the pixel rate is high, if the bit rate is low, the parallel number of variable length coding / decoding processing units may be low.
- the image processing unit changes the parallel number according to the resolution and the frame rate
- the variable length coding / decoding processing unit changes the parallel number according to the bit rate of the stream
- CAVLC is a variable length coding system based on a variable length coding table.
- a code match is compared with a plurality of tables provided in advance, so that processing of several bits to several tens of bits can be performed in one cycle.
- CABAC CABAC-based on a variable length coding table. Therefore, the parallelism of the variable length coding / decoding processing unit may be controlled using this in accordance with CABAC or CAVLC.
- the variable length coding / decoding controller 123 performs control to select two parallel processing in CABAC, in the case of CAVLC, by performing control to perform one parallel processing, It is possible to suppress the operation of unnecessary circuits in the case of CAVLC and perform power consumption control with higher accuracy.
- the image compression / decompression apparatus can select an optimal encoding method according to power consumption.
- H.264 has been described. Although described based on the image compression / decompression device according to H.264, the hybrid encoding method based on frequency conversion processing such as MPEG1, 2, 4 and the like using DCT, run-length encoding, and variable-length encoding processing has similar characteristics, The method of controlling the parallelism independently between the frequency processing unit and the variable length coding / decoding block, which is a feature of the present invention, is effective.
- the image compression / decompression circuit that performs image compression and decompression with a common circuit has been described.
- the image compression apparatus that performs only image compression or the image decompression apparatus that performs only image decompression has a separate configuration. Even so, it has the same effect.
- each buffer in the middle was stored in the external memory.
- SRAM Static Random Access Memory
- image compression / decompression processing circuits that handle various resolutions and bit rates have the same effect on still images represented by JPEG (Joint Photographic Experts Group). .
- JPEG Joint Photographic Experts Group
- DESCRIPTION OF SYMBOLS 100 ... Image compression / expansion apparatus 101 ... Image processing part 1 102: Image processing unit 2 103. Image processing unit 3 104 Image processing unit 4 105 ... External memory 106 ... Original image buffer unit 107 ... Inter search unit 108 ... Intra prediction mode determination unit 109 ... Intra prediction unit 110 ... Residual calculation unit 111 ... Discrete cosine transform unit 112 ... Quantization unit 113 ... Zigzag scan unit 114 ... Run-length encoding unit 115 ... Run-length decoding unit 116 ... Zigzag scanning unit 117 ... Inverse quantization unit 118 ... Inverse discrete cosine transform unit 119 ... Residual calculation unit 120 ...
- Deblocking filter 121 External memory communication bus unit 122 ... Image processing unit controller 123... Image processing unit clock control unit 124... Variable length encoding / decoding processing unit 1 125... Variable length encoding / decoding processor 2 126 ... Variable length encoding unit 127 ... Code buffer unit 128 ... Variable length decoding unit 129 ... Stream combination analysis unit 130 ... Variable length encoding / decoding processing controller 132 ... Parameter input unit (resolution / frame rate) 133 ... Parameter input section (bit rate)
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- Theoretical Computer Science (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
Description
先ず、図1を用いて、一実施形態に係る画像圧縮伸長装置の構成について説明する。図1は、画像圧縮伸長装置の構成の全体構成図である。
次に、図1を用いて、本実施形態に係る画像圧縮伸長装置の各部の構成とその動作について説明する。先ず、画像圧縮伸長装置の画像圧縮処理(以下、単に「圧縮処理」ということもある)について説明する。
上記のように、本実施形態の画像圧縮伸長装置は、複数の画像処理部と、可変長符号化復号処理部を有しており、それらが独立して並列に動作することを前提としている。そのため、次に、図2を用いて画像処理部101、102、103、104と可変長符号化復号処理部124、125の並列数について説明する。図2は、画像処理部101、102、103、104と可変長符号化復号処理部124、125の並列数を決めるモードについて説明する図である。
以下、図1の画像圧縮伸長装置100が、図2のモードNo.1で示した、画像処理部並列数1、可変長符号化復号処理部並列数1で動作する画像圧縮処理、画像伸長処理について説明する。先ず、図3を用いて画像処理部の並列数が1のときの画像(1フレーム)とマクロブロックの関係について説明する。図3は、画像処理部の並列数が1のときの画像(1フレーム)とマクロブロックの関係について説明する図である。
図4は、画像処理部101が、処理ステージごとに、時間単位でマクロブロックを処理する様子を示す図である。図4において、原画像読み出しステージは、原画バッファ部106、イントラ予測モード判定ステージは、イントラ予測モード判定部108、イントラインター判定ステージは、イントラ予測部109、及び、探索部107、予測画作成ステージは、残差計算部110、周波数変換ステージは、離散コサイン変換部111、量子化112、逆量子化117、及び、逆離散コサイン変換部118、ランレングスステージは、ジグザグスキャン部113及び、ランレングス符号化部114、復号画作成ステージは、残差計算部119、デブロッキングフィルタステージは、デブロッキングフィルタ120が、各々動作するステージである。
以下、図1の画像圧縮伸長装置100が、図2のモードNo.5で示した、画像処理部並列数4、可変長符号化復号処理部並列数2で動作する圧縮処理、画像伸長処理について説明する。先ず、図7を用いて、画像処理部の並列数が4のときの画像(1フレーム)とマクロブロックの関係について説明する。図7は、画像処理部の並列数が4のときの画像(1フレーム)とマクロブロックの関係について説明する図である。
次に、図8を用いて、画像処理部101、102、103、104の4系統動作時のマクロブロックのパイプライン処理の様子について説明する。図8は、画像処理部101、102、103、104が、処理ステージごとに、時間単位でマクロブロックを処理する様子を示す図である。
次に、図11及び図12を用いて、画像圧縮伸長装置の画像圧縮処理の詳細手順を説明する。図11は、画像圧縮伸長装置の画像圧縮処理の詳細手順を示すフローチャートである。図12は、画像圧縮処理において、各ピクチャと各部の処理のタイミングを示すタイミングチャートである。
次に、ストリーム結合解析ブロック部129の詳細について説明する。ストリーム結合解析ブロック部129は、例えば、可変長符号化復号処理部124、125が並列動作するときに機能する部分であり、圧縮処理では、各々可変長符号化復号処理部124、125が生成した1フレーム単位のストリームデータの結合を行う。
次に、図15A及び図15Bを用いて、クロック制御部123、131を説明する。図15Aは、クロック制御により、画像処理部と可変長符号化復号処理部の並列処理を制御する例を示す図である。図15Bは、電源制御により、画像処理部と可変長符号化復号処理部の並列処理を制御する例を示す図である。
上記したように、画像処理部は、各パイプライン内の処理は、マクロブロック内の各画素数に依存した処理時間となるため、動作する際の解像度×フレームレートで算出されるピクセルレートに対応して、必要処理スピードを規定できる。一方、可変長符号化復号処理部の処理は、ピクセルレートよりもビットストリームのビットレートに依存してその処理時間が変わる。そのため、画像処理部は、解像度×フレームレート=ピクセルレートに基づいて、並列数を決定し、変長符号化復号処理部は、ビットレートに基づいて並列数を決定するというのが本発明の発想であった。
以上、本実施形態では、H.264による画像圧縮伸長装置に基づいて記載したが、MPEG1、2、4などDCTによる周波数変換処理、ランレングス符号化、可変長符号化処理に基づくハイブリッド符号化方式においては、同様の特性を持ち、本発明の特徴である周波数処理部と可変長符号化復号化ブロック間で独立に並列度を制御する方法が有効である。
以上のように、本発明の好ましい実施形態によれば、システム要件に応じて最適な電力で画像圧縮伸長処理動作を実現することができ、低消費電力化が可能となる。
101…画像処理部1
102…画像処理部2
103…画像処理部3
104…画像処理部4
105…外部メモリ
106…原画バッファ部
107…インター探索部
108…イントラ予測モード判定部
109…イントラ予測部
110…残差計算部
111…離散コサイン変換部
112…量子化部
113…ジグザグスキャン部
114…ランレングス符号化部
115…ランレングス復号部
116…ジグザグスキャン部
117…逆量子化部
118…逆離散コサイン変換部
119…残差計算部
120…デブロッキングフィルタ
121…外部メモリ通信用バス部
122…画像処理部コントローラ
123…画像処理部クロック制御部
124…可変長符号化復号処理部1
125…可変長符号化復号処理部2
126…可変長符号化部
127…符号バッファ部
128…可変長復号部
129…ストリーム結合解析部
130…可変長符号化復号処理コントローラ
132…パラメータ入力部(解像度・フレームレート)
133…パラメータ入力部(ビットレート)
Claims (9)
- 原画像データを符号化に基づいて圧縮し、また、符号化されたデータを伸長する画像圧縮伸長処理を行う装置であって、
イントラ予測又はインター予測による予測画像生成、差分画像生成、周波数変換、量子化・逆量子化を行う一つ以上の画像処理部と、可変長符号化復号処理を行う一つ以上の可変長符号化復号処理部と有し、
画像の解像度又はフレームレートに基づいて、第一の並列数を決定し、前記一つ以上の画像処理部は、前記第一の並列数に応じた並列度により並列処理することを特徴とする画像圧縮伸長装置。 - 前記画像圧縮伸長装置が処理するビットストリームのビットレートに基づいて、第二の並列数を決定し、前記一つ以上の可変長符号化復号処理部は、前記第二の並列数に応じた並列度により並列処理することを特徴とする請求項1記載の画像圧縮伸長装置。
- 前記一つ以上の画像処理部は、前記第一の並列数を決定するために、前記画像の解像度、前記フレームレートの両方、あるいは前記画像の解像度、前記フレームレートの一方をパラメータとして入力されることを特徴とする請求項1記載の画像圧縮伸長装置。
- 前記第一の並列数又は前記第二の並列数に基づき、動作しなくなった前記画像処理部、前記可変長符号化復号処理部のクロックツリーへのクロック供給を止めることを特徴とする請求項2記載の画像圧縮伸長装置。
- 前記第一の並列数又は前記第二の並列数に基づき、動作しなくなった前記画像処理部、前記可変長符号化復号処理部の電源供給を止めることを特徴とする請求項1記載の画像圧縮伸長装置。
- 画像処理部を少なくても二つ以上有し,同一フレーム中の異なるマクロブロックラインを各々の画像処理部で並列処理することを特徴とする請求項1記載の画像圧縮伸長装置。
- 可変長符号化復号処理部を少なくても二つ以上有し,各々の可変長符号化復号処理部では異なるフレームを並列に動作させることを特徴とする請求項2記載の画像圧縮伸長装置。
- 前記第二の並列数を、扱う符号化方式に基づいて決定することを特徴とする請求項2記載の画像圧縮伸長装置。
- 前記符号化方式は、CAVLC及びCABACであることを特徴とする請求項2記載の画像圧縮伸長装置。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2016507303A JP6085065B2 (ja) | 2015-02-09 | 2015-02-09 | 画像圧縮伸長装置 |
US14/913,723 US20160301945A1 (en) | 2015-02-09 | 2015-02-09 | Image compression/decompression device |
CN201580038688.5A CN106576168A (zh) | 2015-02-09 | 2015-02-09 | 图像压缩解压缩装置 |
EP15874402.9A EP3258691A4 (en) | 2015-02-09 | 2015-02-09 | Image compression/decompression device |
PCT/JP2015/053501 WO2016129031A1 (ja) | 2015-02-09 | 2015-02-09 | 画像圧縮伸長装置 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2015/053501 WO2016129031A1 (ja) | 2015-02-09 | 2015-02-09 | 画像圧縮伸長装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016129031A1 true WO2016129031A1 (ja) | 2016-08-18 |
Family
ID=56614479
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/053501 WO2016129031A1 (ja) | 2015-02-09 | 2015-02-09 | 画像圧縮伸長装置 |
Country Status (5)
Country | Link |
---|---|
US (1) | US20160301945A1 (ja) |
EP (1) | EP3258691A4 (ja) |
JP (1) | JP6085065B2 (ja) |
CN (1) | CN106576168A (ja) |
WO (1) | WO2016129031A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018117308A (ja) * | 2017-01-20 | 2018-07-26 | キヤノン株式会社 | 再生装置及びその制御方法 |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3381190B1 (en) * | 2016-08-04 | 2021-06-02 | SZ DJI Technology Co., Ltd. | Parallel video encoding |
CN109429070A (zh) * | 2017-08-23 | 2019-03-05 | 杭州海康威视系统技术有限公司 | 一种移动终端视频编码方法、装置及移动终端 |
CN107483952A (zh) * | 2017-08-29 | 2017-12-15 | 郑州云海信息技术有限公司 | 一种jpeg图像解压缩的方法、装置与系统 |
TW201939953A (zh) * | 2018-03-16 | 2019-10-01 | 晨星半導體股份有限公司 | 影像壓縮系統及利用影像壓縮系統以壓縮影像畫面的方法 |
CN109788299A (zh) * | 2019-03-12 | 2019-05-21 | 广东工业大学 | 一种移动设备间视频传输方法、系统、设备及计算机介质 |
US11776093B2 (en) * | 2019-07-16 | 2023-10-03 | University Of Florida Research Foundation, Incorporated | Automatic sharpness adjustment for imaging modalities |
CN111768460B (zh) * | 2020-06-30 | 2023-08-01 | 北京百度网讯科技有限公司 | 一种解码方法、装置、设备以及存储介质 |
US20230140628A1 (en) * | 2021-11-04 | 2023-05-04 | Meta Platforms, Inc. | Novel buffer format for a two-stage video encoding process |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008067026A (ja) * | 2006-09-07 | 2008-03-21 | Fujitsu Ltd | Mpegデコーダ及びmpegエンコーダ |
JP2008141531A (ja) | 2006-12-01 | 2008-06-19 | Canon Inc | 画像符号化装置及び画像符号化方法 |
JP2010109572A (ja) * | 2008-10-29 | 2010-05-13 | Toshiba Corp | 画像処理装置、及び方法 |
JP2011160077A (ja) * | 2010-01-29 | 2011-08-18 | Sony Corp | 復号装置および方法 |
WO2012014471A1 (ja) * | 2010-07-30 | 2012-02-02 | パナソニック株式会社 | 画像復号化装置、画像復号化方法、画像符号化装置および画像符号化方法 |
WO2012160879A1 (ja) * | 2011-05-20 | 2012-11-29 | ソニー株式会社 | 画像符号化装置と画像復号装置およびその方法 |
JP2014096755A (ja) * | 2012-11-12 | 2014-05-22 | Canon Inc | 画像符号化装置、画像符号化方法及びプログラム、画像復号装置、画像復号方法及びプログラム |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1157641C (zh) * | 1997-09-03 | 2004-07-14 | 松下电器产业株式会社 | 处理器 |
JP4699685B2 (ja) * | 2003-08-21 | 2011-06-15 | パナソニック株式会社 | 信号処理装置及びそれを用いた電子機器 |
JPWO2010125616A1 (ja) * | 2009-04-27 | 2012-10-25 | パナソニック株式会社 | 符号化ストリーム復号装置 |
US10244246B2 (en) * | 2012-02-02 | 2019-03-26 | Texas Instruments Incorporated | Sub-pictures for pixel rate balancing on multi-core platforms |
US9426498B2 (en) * | 2012-07-10 | 2016-08-23 | Broadcom Corporation | Real-time encoding system of multiple spatially scaled video based on shared video coding information |
US20140205012A1 (en) * | 2013-01-21 | 2014-07-24 | Mediatek Inc. | Method and apparatus using software engine and hardware engine collaborated with each other to achieve hybrid video encoding |
US9596477B2 (en) * | 2013-12-26 | 2017-03-14 | Mediatek Inc. | Methods of multiple-slice coding for frame buffer compression |
CA2950171C (en) * | 2014-05-28 | 2019-11-19 | Arris Enterprises Llc | Content aware scheduling in a hevc decoder operating on a multi-core processor platform |
US9965021B2 (en) * | 2014-06-23 | 2018-05-08 | Mediatek, Inc. | Method and system providing power management for multimedia processing |
-
2015
- 2015-02-09 EP EP15874402.9A patent/EP3258691A4/en not_active Withdrawn
- 2015-02-09 JP JP2016507303A patent/JP6085065B2/ja active Active
- 2015-02-09 WO PCT/JP2015/053501 patent/WO2016129031A1/ja active Application Filing
- 2015-02-09 CN CN201580038688.5A patent/CN106576168A/zh active Pending
- 2015-02-09 US US14/913,723 patent/US20160301945A1/en not_active Abandoned
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008067026A (ja) * | 2006-09-07 | 2008-03-21 | Fujitsu Ltd | Mpegデコーダ及びmpegエンコーダ |
JP2008141531A (ja) | 2006-12-01 | 2008-06-19 | Canon Inc | 画像符号化装置及び画像符号化方法 |
JP2010109572A (ja) * | 2008-10-29 | 2010-05-13 | Toshiba Corp | 画像処理装置、及び方法 |
JP2011160077A (ja) * | 2010-01-29 | 2011-08-18 | Sony Corp | 復号装置および方法 |
WO2012014471A1 (ja) * | 2010-07-30 | 2012-02-02 | パナソニック株式会社 | 画像復号化装置、画像復号化方法、画像符号化装置および画像符号化方法 |
WO2012160879A1 (ja) * | 2011-05-20 | 2012-11-29 | ソニー株式会社 | 画像符号化装置と画像復号装置およびその方法 |
JP2014096755A (ja) * | 2012-11-12 | 2014-05-22 | Canon Inc | 画像符号化装置、画像符号化方法及びプログラム、画像復号装置、画像復号方法及びプログラム |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018117308A (ja) * | 2017-01-20 | 2018-07-26 | キヤノン株式会社 | 再生装置及びその制御方法 |
JP7020782B2 (ja) | 2017-01-20 | 2022-02-16 | キヤノン株式会社 | 再生装置及びその制御方法 |
Also Published As
Publication number | Publication date |
---|---|
EP3258691A1 (en) | 2017-12-20 |
US20160301945A1 (en) | 2016-10-13 |
CN106576168A (zh) | 2017-04-19 |
EP3258691A4 (en) | 2018-10-31 |
JPWO2016129031A1 (ja) | 2017-04-27 |
JP6085065B2 (ja) | 2017-02-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6085065B2 (ja) | 画像圧縮伸長装置 | |
US11647190B2 (en) | Image coding apparatus, image coding method, image decoding apparatus, image decoding method, and program | |
KR101156997B1 (ko) | 부호화 장치 및 부호화 방법 | |
US20060115002A1 (en) | Pipelined deblocking filter | |
US9270988B2 (en) | Method of determining binary codewords for transform coefficients | |
US20080240587A1 (en) | Selective information handling for video processing | |
JP2006157925A (ja) | パイプラインデブロッキングフィルタ | |
JPH10271516A (ja) | 圧縮符号化装置、符号化方法、復号化装置および復号化方法 | |
JP2007259323A (ja) | 画像復号化装置 | |
WO2016047375A1 (ja) | 動画像符号化装置、動画像復号装置、および動画像符号化・復号化方法 | |
JP2008289105A (ja) | 画像処理装置およびそれを搭載した撮像装置 | |
JP4880868B2 (ja) | 性能向上のための動画像圧縮方法及び装置 | |
JP7451131B2 (ja) | 画像符号化装置、画像符号化方法、及びプログラム | |
WO2017104010A1 (ja) | 動画像符号化装置および動画像符号化方法 | |
JP2010268094A (ja) | 画像復号化装置および画像復号化方法 | |
WO2016143093A1 (ja) | 動画像符号化装置及びそれに用いるIntra予測符号化方法、及び動画像復号化装置 | |
JPH1127667A (ja) | 画像符号化装置および画像復号装置 | |
JP2001268571A (ja) | 動画像再符号化装置 | |
JP2016005116A (ja) | 画像符号化装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2016507303 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14913723 Country of ref document: US |
|
REEP | Request for entry into the european phase |
Ref document number: 2015874402 Country of ref document: EP |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15874402 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |