WO2007132539A1 - 画像圧縮装置、圧縮方法及びプログラム並びに画像復元装置、復元方法及びプログラム - Google Patents
画像圧縮装置、圧縮方法及びプログラム並びに画像復元装置、復元方法及びプログラム Download PDFInfo
- Publication number
- WO2007132539A1 WO2007132539A1 PCT/JP2006/316107 JP2006316107W WO2007132539A1 WO 2007132539 A1 WO2007132539 A1 WO 2007132539A1 JP 2006316107 W JP2006316107 W JP 2006316107W WO 2007132539 A1 WO2007132539 A1 WO 2007132539A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- quantization
- pixel
- control signal
- quantization table
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
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
- H04N19/90—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
-
- 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
- H04N19/124—Quantisation
-
- 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
- H04N19/124—Quantisation
- H04N19/126—Details of normalisation or weighting functions, e.g. normalisation matrices or variable uniform quantisers
-
- 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
- H04N19/14—Coding unit complexity, e.g. amount of activity or edge presence estimation
-
- 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
- H04N19/182—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 a pixel
-
- 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
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
Definitions
- Image compression apparatus compression method and program
- image restoration apparatus restoration method and program
- the present invention provides a high-quality real-time image compression apparatus, compression method, and compression effective for both two types of images having different characteristics (natural images such as movies and CG images such as digital maps).
- the present invention relates to a program for performing image restoration, an image restoration apparatus for restoring compressed image data, a restoration method, and a program for performing the restoration.
- data compression In order to transmit image information including a huge amount of information, data compression is necessary. In particular, in the case of moving images, data compression is indispensable because 30 to 60 frames of image data are transmitted and received per second.
- a first conventional technique using a DPCM predictor as shown in FIG. 1 is known.
- the predictor (DPCM) 210 predicts the level value of the prediction target pixel from the previous line of the target pixel and the value of the previous pixel, and the actual pixel level.
- the prediction error with the value is converted into a representative value, quantized by the quantizer 310, and transmitted via the variable length encoder 410 to which a code corresponding to the appearance frequency is assigned.
- quantization and coding are performed on a pixel-by-pixel basis, it can be applied regardless of high frequency or low frequency.
- the prediction value is calculated based on the value of the previous line, if a prediction error occurs at any point in time, the error is used as it is in the next prediction, and as a result, the prediction error propagates. End up along the line There is a problem that quality deterioration occurs.
- the quantization table used by the quantizer 310 is a quantization table in which the prediction error level value is 4 to 4 and the quantization prediction error is 0 (that is, the quantization width is coarse).
- the quantization prediction error is regarded as zero. For this reason, the information that the prediction error has occurred cannot be included in the quantization result, and as a result, the prediction error remains as it is.
- lines in the line direction that should not have existed in the image that is output as a processing result leading to degradation of image quality.
- a second conventional technique is known in which data compression is performed by JPEGOoint Photographic Experts Group (MPEG) and Moving Picture Experts Group (MPEG) as shown in FIG.
- MPEG Joint Photographic Experts Group
- MPEG Moving Picture Experts Group
- Patent documents 1 and 2 listed below can be cited as patent documents belonging to the second prior art.
- the second conventional technique blocks the input image 120 (usually 8 * 8 pixels) 130, applies DCT 220 to the block image, and applies quantization 320 to the DCT coefficient. Then, the variable length coding 420 to which a code corresponding to the appearance frequency is assigned is transmitted.
- DCT Discrete Cosine Transfer
- conversion is a technique for frequency conversion of image data.
- the human eye Since the human eye is sensitive to low-frequency components (flat areas in the image), the DCT coefficients related to low frequencies are fine, and the DCT coefficients related to high frequencies are coarsely quantized. It is possible to compress at a high compression rate as it is conspicuous. However, there is no problem with compression of natural images because the low-frequency components that are easily noticeable by humans are quantized, but there is no problem with high-frequency components such as lines and characters in map images (CG images). Stands out. Furthermore, since edge information extraction 230 is performed on the block to be compressed, there is a problem that image quality deterioration due to quantization error can be found, corrected, and fed back.
- a third conventional technique using JPEG-LS which is not shown, is known.
- the third conventional technique performs level value prediction with a MED (Median Edge Detector) predictor and directly codes the prediction error.
- MED Median Edge Detector
- no quantization is performed, so image quality does not deteriorate. But the mark There is a problem that each process is not suitable for real-time compression, such as requiring calculation processing at the time of issue.
- Patent Document 1 JP-A-10-126777
- Patent Document 2 Japanese Patent Application Laid-Open No. 6-350092
- an object of the present invention is to change the quantization table in consideration of the continuity of the prediction level values between the peripheral pixels and the lines, and to generate lines caused by quantization errors.
- An image compression apparatus, a compression method, a program for performing the compression, an image restoration apparatus for restoring compressed image data, a restoration method, and a program for performing the restoration are provided.
- An image compression apparatus includes a predictor that predicts a pixel level value of a compression target pixel.
- the image compression apparatus includes a line (line) to which the compression target pixel belongs and a preceding line.
- Specific pixel level difference evaluation means for evaluating the identity of the level difference between specific pixels of the compressed pixels relating to the line and sending a control signal to the quantization table switching means, and
- absolute value sum evaluation means for evaluating the sum of absolute values of pixel level differences around the compression target pixel and sending a control signal to the quantization table switching means.
- the quantization table switching unit instructs the switching of the quantization table and has a plurality of quantization steps having different quantization steps. Select one of the conversion tables And quantize.
- image quality degradation is caused by evaluating whether image quality degradation has occurred in the quantization process on a pixel-by-pixel basis and immediately feeding it back to the quantization and sign of the next pixel. In this case, correction can be made instantaneously (in units of several pixels).
- the quantization step can be coarsened or reduced in strength. In other words, it is possible to increase the compression efficiency while preventing image quality deterioration.
- the image compression method of the present invention and the program for compressing the image of the present invention are a method and a program executed in an image compression apparatus including a predictor that predicts a pixel level value of a compression target pixel.
- a control signal that determines the presence or absence of a quantization error based on how the same value appears in the level difference between specific pixels between two rows and instructs the quantizer to switch the quantization table used.
- image quality degradation is caused by evaluating whether image quality degradation has occurred in the quantization process on a pixel-by-pixel basis and immediately feeding back to the quantization and sign of the next pixel. In this case, correction can be made instantaneously (in units of several pixels).
- quantization error occurs, the same level difference for the quantization step often occurs between the rows, so the level difference of the same value is evaluated and the quantization error propagates. Propagation of prediction error that occurs from the vicinity of the edge toward a plane with the same level difference as a CG image by performing immediate feedback and quantization when it is determined that it is V. It becomes possible to prevent.
- the quantization step can be made coarser or less powerful. In other words, it is possible to increase the compression efficiency while preventing image quality deterioration.
- the image restoration apparatus and the image restoration method of the present invention restore the image data compressed by the above-described image compression apparatus by a reverse operation, is the image quality degraded during the inverse quantization process? Is evaluated in units of pixels, and is immediately fed back to the inverse quantization and inverse sign of the next pixel. If image quality degradation occurs, it is corrected instantaneously (in units of several pixels). It becomes possible.
- the quantization step can be roughened or reduced in strength. In other words, restoration efficiency can be improved while preventing image quality degradation.
- the program for restoring image data according to the present invention is an operation reverse to the operation by the program for compressing image data, and is similar to the program for compressing image data described above.
- the specific pixel level difference evaluation process or the absolute value sum evaluation process can be performed at high speed
- the inverse quantization table switching process can be performed at high speed.
- FIG. 1 is a diagram showing an outline of a first conventional technique using a DPCM predictor.
- FIG. 2 is a diagram showing an outline of a second prior art that performs data compression by JPEG and MPEG.
- FIG. 3 is an operation flowchart of an image compression method according to the principle configuration of the present invention.
- FIG. 4 is a block diagram showing a configuration of an image data compression (encoding) apparatus according to an embodiment of the present invention.
- FIG. 5 is a diagram showing an arrangement relationship between a compression target pixel and peripheral pixels in various image formats according to the embodiment of the present invention.
- [6A] This is a quantization table that represents the correspondence between the quantization value and the quantization number for the prediction error ( ⁇ — ⁇ ′), and is the first quantization table that also has a fine quantization step force.
- [6]] This is a quantization table that represents the correspondence between the quantization value and the quantization number for the prediction error ( ⁇ — ⁇ '), and is the second quantization table that has a normal quantization step force.
- [6C] This is a quantization table that represents the correspondence between the quantization value and the quantization number for the prediction error ( ⁇ — ⁇ ′), and is a third quantization table that also includes coarse and quantum stepping forces.
- FIG. 7 is a flowchart for explaining the operation of the continuity evaluation result feedback module shown in FIG. 4.
- FIG. 8 is a flowchart for explaining the operation of the peripheral level difference determination module shown in FIG.
- FIG. 9 is a flowchart for explaining the operation of the image data compression (encoding) apparatus according to the embodiment of the present invention.
- FIG. 9B A flowchart illustrating the processing of the “local decoder” in step S008 in FIG. 9B.
- FIG. 10 is a functional block diagram showing a system configuration of the image data compression apparatus according to the embodiment of the present invention.
- FIG. 11 is a block diagram showing the configuration of the image data restoration (decoding) device according to the embodiment of the present invention.
- FIG. 12C An inverse quantization table representing the association of the quantization prediction error with the quantization number provided in the inverse quantizer 033, and then a third inverse quantization table that also has a quantization step force. is there.
- FIG. 13 is a flowchart for explaining the operation of the continuity evaluation result feedback module shown in FIG. 11.
- FIG. 14 is a flowchart for explaining the operation of the peripheral level difference determination module shown in FIG. 11.
- FIG. 15A is a flowchart for explaining the operation of the image data restoration (decoding) device according to the embodiment of the present invention.
- FIG. 15B is a flowchart for explaining the processing of the “local decoder” in step S028 in FIG. 15A.
- FIG. 16 is a functional block diagram showing the system configuration of the image data restoration apparatus according to the embodiment of the present invention.
- FIG. 3 is an operation flowchart of the image compression method according to the principle configuration of the present invention.
- the image compression process in the principle configuration of the present invention shown in FIG. 3 is greatly different from the configuration of the first prior art described above.
- the image compression process includes an adaptive quantizer 300, and the level value of the prediction target pixel X and the predictor 200 are When the prediction error from the predicted value X ′ is quantized by the quantizer, the adaptive quantizer 300 uses the continuity of the level value of the pixel of interest between the lines or the level difference measuring power of the surrounding pixels. This is the point of adaptively selecting quantization tables with different steps. Then, quantization is performed based on an adaptively selected quantization table, and transmitted through a variable length encoder 400 to which a code corresponding to the appearance frequency is assigned.
- FIG. 4 is a block diagram showing a configuration of the image data compression (encoding) apparatus according to the embodiment of the present invention.
- image data 001 represents a group of pixels to be compressed, and is extracted by an image processing apparatus (not shown) that digitally processes an image.
- a compression target pixel X (002) represents a pixel to be compressed in the current step.
- Figure 5 shows an image frame.
- FIG. 5 is a diagram showing an arrangement relationship between a compression target pixel X and peripheral pixels (between lines and previous pixels) when the format is an interlaced image or a progressive image. In Figure 5! / The line is indicated by a broken line.
- the quantizer 003 measures the continuity of the level values of the target pixel (pixels A and B in this example) between the lines, the level difference measurement of peripheral pixels (the sum of the absolute values of the level differences of the peripheral pixels in this example). ) To select a quantization table with a different quantum step from the quantization table switching module 017 according to the quantization table selected from the control signal 017 and the predicted value X ′ (006 ), Which is the difference value of), the prediction error (X—X ') is input, and the prediction error quantization value (also simply referred to as the quantization value) and the quantization number are Output.
- FIG. 6A is a quantization table showing a correspondence between a quantization value and a quantization number for a prediction error ( ⁇ ′), and is a first quantization table having a fine quantization step force.
- Fig. 6 (b) is a quantization table that represents the correspondence between the quantization value and the quantization number for the prediction error (X – X '), and is the second quantization table that has a normal quantization step force.
- FIG. 6C is a quantization table that represents the correspondence between the quantization value and the quantization number for the prediction error (X ⁇ X ′), and is a third quantization table that also has a quantization step force.
- the width of the quantization step is preferably narrower as the absolute value of the prediction error is smaller, and is wider as the absolute value of the prediction error is larger.
- This is generally a prediction error force, and in this case, it is a flat image that is easy to predict, and conversely, if the prediction error is large, it tends to be near the edge where the prediction is easily lost, and there is an error in the flat part. This is to reduce the quantization error by narrowing the step width when the prediction error is small.
- the minimum step width is preferably about 1-4.
- the quantization step width is widened to increase the compression efficiency.
- This step width is preferably 24-32.
- the portion near the edge where the appearance frequency is low is basically less noticeable than the case where there is an error in the flat portion even if the prediction error is large. Must not.
- the step width is increased to increase the compression efficiency, and on the contrary, the flat portion is further finely quantized.
- the absolute value of the prediction error is set to a minimum of 0 and set to a maximum of 11 or more.
- the absolute value of the prediction error is set to a minimum of 2, and set to a maximum of 21 or more.
- the absolute value of the prediction error is set to a minimum of 4, and set to a maximum of 41 or more.
- the decompression side has an inverse quantization table (see FIGS. 12A, 12B, and 12C) indicating the correspondence between the quantization value and the quantization number corresponding to FIGS. 6A, 6B, and 6C. This will be described later. Also, in the above description, the power shown in the three examples of quantization tables is not limited to this, but there are two or more quantization tables having different quantum step widths.
- the predetermined compression ratio may not be reached. Therefore, a plurality of quantization tables having different quantization step widths are provided, and the predetermined compression ratio is set. In such a case, adaptive quantization may be performed in which a quantization table with a coarse step width is selected.
- the variable length encoder 004 receives the quantization number obtained from any one of the quantization tables in Figs. 6A, 6B, and 6C and outputs a variable length code.
- the compression code buffer 005 is a buffer for accumulating the output of the variable length encoder 004.
- the predicted value 006 is a predicted value calculated by the predictor 018.
- the peripheral pixel A (007) is a pixel on the left of the compression target pixel X as shown in FIG.
- the one-line buffer for prediction 008 is a buffer that holds about one line of the quantization result for prediction, and is constituted by a shift register, for example.
- the peripheral pixel D (009) is a pixel at the upper right of the compression target pixel X, that is, a pixel in the next column in the line processed immediately before.
- the peripheral pixel C (010) is a pixel above the compression target pixel X, that is, a pixel in the same column in the previous processed line.
- the peripheral pixel B (011) is the upper left pixel of the compression target pixel X, that is, the pixel in the previous column in the previous processed line.
- the peripheral pixel E (012) is the pixel on the left of the peripheral pixel B (011), that is, the pixel in the second previous column in the previously processed line.
- peripheral pixel D peripheral pixel C, peripheral pixel B, and peripheral pixel E corresponds to a pixel on compression line X that is two lines above in the case of an interlaced image, and one line above in the case of a progressive image. This is the pixel.
- the continuity evaluation result feedback module 013 measures the level difference based on the surrounding pixels A (007) and B (011), evaluates the continuity, and repeats the same level difference more than the specified number of times.
- the control signal 014 is output so that a fine quantization table (first quantization table shown in FIG. 6A) is used.
- FIG. 7 is a flowchart for explaining the operation of the continuity evaluation result feedback module 013 shown in FIG.
- step S041 in Fig. 7 the level difference B—A is compared with the previous level difference.
- step S042 it is determined whether the level difference B ⁇ A is the same as the previous level difference. If not, the same value continuous counter is set to 1 in step S043 and the process proceeds to step S045. If they are the same, the process proceeds to step S044, and 1 is added to the same value continuous counter in step S044.
- step S045 it is determined whether the same value continuous counter is 4 or more. If the value of the counter is 4 or more, the process proceeds to step S046, and in step S046, the control signal “000” indicating the quantization table of the fine quantization step is transmitted and the process ends. In S045, if it is greater than the value of the counter, the level value of the same value has continued for a predetermined number of times, and it is suspected that the prediction error has propagated. By switching to the step-step quantization table, the error of the prediction level value can be detected more finely. Therefore, when propagation of prediction error has actually occurred, it is possible to prevent further propagation of the prediction error by switching the quantization table by this processing.
- step S047 control is transferred to the peripheral level difference detection module 015 and the process is terminated.
- control signal 014 is fine, the control signal “000” indicating the quantization step, or the peripheral level difference detection module 015 is activated to determine which quantization table to use again.
- the control signal to show is shown.
- the peripheral level difference determination module 015 determines the quantization table to be used by calculating the absolute value of the sum of the level differences of the peripheral pixels and comparing it with a threshold value. After the determination, control signal 016 is transmitted. That is, the control signal 016 transmits the following control signal.
- Control signal 000 fine, transmitted when using quantization step quantization table
- Control signal 001 transmitted when using the quantization table of the normal quantization step
- Control signal 010: 3 ⁇ 4 transmitted when using the quantization table of the quantization step
- FIG. 8 is a flowchart for explaining the operation of the peripheral level difference determination module 015 shown in FIG.
- Ndsub which is the sum of absolute values of level difference D—C, level difference C—B, level difference B—A, and level difference B—E is calculated.
- step S052 it is determined whether Ndsub is 15 or less. If Ndsub is 15 or less, the process proceeds to step S053, and the control signal “000” indicating the quantization table of the quantization step is transmitted to step S053 and the process is terminated.
- a low-activity image flat image
- a fine quantization table is used for such a low-activity image. By doing so, high-precision encoding is performed.
- the threshold value of 15 is used in the above description, but is not limited to this.
- the pixel level difference is about 2 to 4, it is easily detected by human eyes. Since the absolute sum of the four level differences is taken here, it is easy for human eyes to detect! /, The total pixel level difference is in the range of 8 (ie 2 X 4) to 16 (ie 4 X 4). Become.
- the threshold used for the above judgment is the error in an image with low activity (flat image) Since this is to suppress propagation, a threshold (15 in the above) should be determined within the range of 8-16 above.
- step S054 it is determined whether or not Ndsub is 128 or more in step S054. If Ndsub is not equal to or greater than 128, the process proceeds to step S055, and the control signal “001” indicating the quantization table of the normal quantization step is transmitted to step S055 and the process is terminated. If Ndsub is equal to or greater than 128, the process proceeds to step S056. In step S056, the control signal “010” indicating the quantization table of the quantization step is transmitted and the process ends. In other words, an image with high activity (an image with few flat parts) is difficult for the human eye to understand even if some image degradation occurs. Use the quantization table to perform high-speed encoding.
- the threshold value of 128 is used in the above description, but is not limited to this.
- the quantum step width of images with high activity is set to 24 to 32, so the absolute sum of the above four level differences is four times that,
- the range is 96-128.
- a threshold value (128 in the above) can be determined, which can be used to distinguish between using the third quantization table for the quantization step and the second quantization table for the normal quantization step. .
- 3-bit control signals (000), (001), (010) are associated as control signals! /, but it is possible to specify quantization tables of different quantization steps. If so, it does not have to be a control signal of this form.
- the quantization table is not limited to the fine, coarse, normal quantization step quantization table as described above, and is further divided into four or more quantization tables, or fine, coarse 2 The only requirement is to provide two or more types of quantization tables with different quantization step widths.
- the quantization table switching module 017 outputs a control signal for switching the quantization table to the quantizer 003 in accordance with the two control signals 014 and 016.
- the predictor 018 calculates a predicted value X of the compression target pixel X based on the surrounding pixels A (007), B (011), and C (010).
- FIG. 9A illustrates the operation of the image data compression (encoding) apparatus according to the embodiment of the present invention. It is a flowchart for doing.
- FIG. 9B is a flowchart for explaining the processing of the “single decoder” in step S008 in FIG. 9A. Steps are abbreviated as S in Figures 9A and 9B.
- S002 When the continuity evaluation result feedback module 013 measures the level difference based on peripheral pixels A (007) and B (011), evaluates continuity, and the same level difference continues for the specified number of times. Outputs a control signal 014 using a fine quantization table to the quantization table switching module 017. Otherwise, go to S003.
- Peripheral level difference detection module 015 detects the sum of the absolute values of the peripheral level differences and evaluates the sum of the absolute values of the peripheral level differences to determine the quantization table. Output to 017.
- the predictor 018 calculates predicted values X and (006) of the compression target pixel X (002).
- the prediction error is input to the quantization table determined by the quantization table switching module 017, and the prediction error quantization value (see FIGS. 6A to 6C) is calculated.
- variable length encoder 004 generates a code based on the quantization number. This code may be a Golomb code or an arithmetic code.
- the peripheral pixel B (011) is substituted as the peripheral pixel E (012) of the next compression target pixel.
- peripheral pixel C (010) is substituted as the peripheral pixel B (011) of the next compression target pixel.
- SOU Assign peripheral pixel D (009) as peripheral pixel C (010) of the next compression target pixel.
- FIG. 10 is a functional block diagram showing the system configuration of the image data compression apparatus according to the embodiment of the present invention. The contents described above are functionalized and expressed in blocks.
- the image data compression apparatus according to the embodiment of the present invention inputs the image data 102 to be compressed to the compression processing unit 110.
- the compression processing unit 110 first inputs the input image data 102 for each line.
- the reading unit 111 reads and extracts the compression target pixels, and inputs the extracted compression target pixels to the prediction processing unit 112.
- the prediction processing unit 112 calculates the prediction error by subtracting the compression target pixel level force from the prediction value of the compression target pixel predicted by the predictor, while the evaluation unit 113 performs the continuity evaluation result of FIG.
- the continuity evaluation unit 114 corresponding to the feedback module 013 evaluates whether the level difference between the two specific peripheral pixels related to the compression target pixel is continuous, and when the same level difference continues for a predetermined number of times, for example, Generate a control signal to use the fine quantization table 119 as shown in Figure 6A.
- the peripheral evaluation unit 115 corresponding to the peripheral level difference detection module 015 in FIG. 4 of the evaluation unit 113 calculates the level difference between two peripheral pixels including pixels around the pixel for which the prediction value of the compression target pixel is calculated.
- the sum of absolute values is detected, and the sum of absolute values of peripheral level differences is evaluated by referring to the level difference table 116 that manages the level difference and the quantization table information to be used in association with each other.
- a control signal indicating which level of the quantization table 119 is used is input to the switching processing unit 117.
- the switching processing unit 117 corresponding to the quantization table switching module 017 instructs the quantization processing unit 118 to switch the quantization table 119 based on the input control signal, and the quantization processing unit 118 performs quantization.
- Table 119 is switched, the prediction error quantization value is calculated using the switched quantization table 119, and is input to the encoding processing unit 120 corresponding to the variable length encoder 004 as the output of the quantization processing unit 118,
- the encoding processing unit 120 generates a compression code based on the input prediction error quantization value, accumulates the generated compression code as an output of the compression processing unit 110, and obtains compressed data 104.
- the processing in the image data compression apparatus shown in FIG. 10 described above can be realized on a computer.
- the hardware resources of the computer Although not shown, various registers, arithmetic units including ALU, RAM, ROM, I / O, etc. are used.
- the above-mentioned modules, predictors, quantizers including quantization tables, encoders, etc. are integrated into a single-chip LSI and used as various image data compression devices for the above-mentioned two types of image data with different characteristics. Is possible.
- the prediction buffer that holds the already quantized pixel level value for one line is used, and quantization and encoding are performed on a pixel basis based on the prediction buffer.
- whether or not image quality degradation has occurred during the quantization process is evaluated on a pixel-by-pixel basis, and immediately fed back to the quantization and sign of the next pixel (to switch the quantization table).
- the quantization error is eliminated (or reduced) by switching the quantization table to a step with a small step width and quantizing it. When deterioration occurs, it can be corrected instantaneously (in units of several pixels).
- the quantization step can be made rougher or more powerful, It is possible to increase the compression efficiency while preventing image quality deterioration.
- the image compression apparatus of the present invention it is detected by a simple device called a counter whether the same level difference appears in a specific pixel continuously, and the same level difference continues for a predetermined number of times or more, for example, 4 times continuously. If control is performed so as to select a quantization table that performs fine quantization when it appears, correction can be made before image quality degradation in the long line direction occurs. [0058] Further, according to the image compression apparatus of the present invention, the sum of the absolute values of the level differences of the peripheral pixels of the compression target pixel is calculated using a prediction buffer that holds the pixel level value for one line, and the sum is calculated. Since the control signal indicating which quantization table is used is sent to the quantization table switching module based on the value of, the quantizer is based on the activity around the pixel to be compressed (pixel level difference between surrounding pixels). Instant feedback is possible.
- the quantization table is switched by the control signal for switching the quantization table, the continuity of the pixel level difference and the activity around the compression target pixel are obtained. It is possible to use three types of quantization tables with different quantization details, that is, fine, normal, and coarse quantization tables. When finer control is performed, it is clear that the same effect can be obtained even if there are more quantization tables than three types.
- the image compression apparatus provides a stream unit in a frame unit when the image to be processed is a progressive image and a stream unit in an order in which the upper part of the screen is directed to the lower unit in the case of an interlaced image.
- the image data flowing in the shape is the processing target.
- the compressed data after compression encoding is packetized and transferred for each fixed size according to the transmission protocol of the transmission line to which the present apparatus is connected.
- FIG. 11 is a block diagram showing the configuration of the image data restoration (decoding) device according to the embodiment of the present invention.
- a compression code 031 represents a compression code of the image data subjected to variable length coding in FIG.
- the inverse encoder 032 receives the compressed code of the variable length encoded image data as an input, and outputs a quantization number corresponding to the code.
- the inverse quantizer 033 receives the quantization number and outputs a prediction error quantized value X (034).
- FIG. 12A is an inverse quantization table showing the correspondence of the quantization value (prediction error quantization value) to the quantization number provided in the inverse quantizer 033, and also includes a detailed quantization step. It is the 1st inverse quantization table which consists of.
- FIG. 12B is an inverse quantization table showing the correspondence of the quantization value (prediction error quantization value) to the quantization number provided in the inverse quantizer 033, and is a second quantization step having a normal quantization step force. This is the inverse quantization table.
- Figure 12C shows the quantization value (prediction for the quantization number provided in the inverse quantizer 033).
- the correspondence between the quantization value and the quantization number shown in FIGS. 12A, 12B, and 12C needs to be the same on the decompression side and the compression side.
- a quantization table (see FIGS. 6A, 6B, and 6C) that indicates the correspondence between the quantization value and the quantization number is provided on the compression side corresponding to FIGS. 12A, 12B, and 12C.
- three examples of inverse quantizers are shown, but the present invention is not limited to this, and two or more may be used.
- the arrangement relationship between the restoration target pixel X and the surrounding pixels is as follows. This is the same as the arrangement relationship with pixels (between lines and previous pixels). However, the compression target pixel in FIG. 5 is read as the restoration target pixel.
- the prediction error quantized value X (034) is added to the prediction value X '(036) calculated by the predictor 048 to obtain restored image data 035, and the surrounding pixel A of the next restoration target pixel is obtained. It becomes (037).
- the peripheral pixel A (037) is a pixel on the left side of the restoration target pixel, like the peripheral pixel A shown in FIG.
- the one-line buffer for prediction 038 is a buffer that holds about one line of the quantization result for prediction, and is composed of, for example, a shift register.
- the peripheral pixel D (039) is the pixel on the upper right (previous line) of the restoration target pixel, and the peripheral pixel C (040) is the restoration target.
- the continuity evaluation result feedback module 043 measures the level difference based on the surrounding pixels A (037) and B (041), evaluates the continuity, and the same level difference continues more than the specified number of times.
- the control signal 044 is output so as to use the fine quantization table (the first inverse quantization table shown in FIG. 12A).
- FIG. 13 is a flowchart for explaining the operation of continuity evaluation result feedback module 043 shown in FIG.
- step S061 in Fig. 13 the level difference B—A is compared with the previous level difference.
- step S062 it is determined whether the level difference B—A is the same as the previous level difference. If they are not the same, 1 is set in the same value continuous counter in step S063 and the process proceeds to step S065. If they are the same, the process proceeds to step S064, and 1 is added to the same value continuous force counter in step S064. In step S065, it is determined whether the same value continuous counter is 4 or more.
- step S066 If it is equal to or greater than the value of the counter, the process proceeds to step S066, and in step S066, a control signal “000” indicating an inverse quantization table of a fine quantization step is transmitted and the process is terminated. If the value is not greater than the value of the counter, the process proceeds to step S067. In step S067, control is transferred to the peripheral level difference detection module 045, and the process ends.
- control signal 044 is fine, the control signal "000" indicating the quantization step, or the peripheral level difference detection module 045 is operated to determine which inverse quantization table to use again.
- the control signal which shows making it show is shown.
- the peripheral level difference determination module 045 calculates the absolute value of the sum of the level differences of the peripheral pixels, and compares this with a threshold value to determine which inverse quantization table to use. After the determination, control signal 046 is transmitted. That is, the control signal 046 transmits the following control signal.
- Control signal 000 fine, transmitted when using a quantization step inverse quantization table
- FIG. 14 shows the peripheral level difference determination module shown in FIG. This is a flowchart explaining the operation of.
- Ndsub which is the sum of absolute values of level difference D—C, level difference C—B, level difference BA, and level difference B—E is calculated.
- step S072 it is determined whether Nds ub is 15 or less. If Ndsub is 15 or less, the process proceeds to step S073, where the control signal “00 0” indicating the inverse quantization table of the quantization step is transmitted to step S073, and the process ends.
- step S074 it is determined in step S074 whether Ndsub is 128 or more. If Ndsub is not 128 or more, the process proceeds to step S075, and in step S075, the control signal “001” indicating the inverse quantization table of the normal quantization step is transmitted and the process is terminated. If Ndsub is equal to or greater than 128, the process proceeds to step S076, and in step S076, the control signal “010” indicating the inverse quantization table of the coarse quantization step is transmitted, and the process ends.
- the threshold value for determining Ndsub is not limited to the values of 15 and 128, and what range of values is desirable V. It is the same as that.
- 3-bit control signals (000), (001), and (010) are associated with each other as control signals! /, but it is possible to indicate an inverse quantization table for different quantization steps. If it is a thing, it does not have to be a control signal of this form.
- the inverse quantization table is not limited to the above-described fine, coarse, normal quantization step inverse quantization table, but may be further divided into four or more inverse quantization tables, or fine, coarse, It is only necessary to provide two types of inverse quantization tables. It is sufficient if two or more types of inverse quantization tables are provided corresponding to the above-mentioned quantization tables on the image data compression apparatus side.
- the quantization table switching module 047 outputs a control signal for switching the inverse quantization table to the inverse quantizer 033 in accordance with the two control signals 044 and 046.
- the predictor 048 calculates the predicted value X ′ of the restoration target pixel X based on the surrounding pixels A (037), B (041), and C (040).
- FIG. 15A is a flowchart for explaining the operation of the image data restoration (decoding) device according to the embodiment of the present invention.
- FIG. 15B is a flowchart for explaining the processing of the “local decoder” in step S028 in FIG. 15A.
- step is abbreviated as S.
- Peripheral level difference detection module 045 detects the sum of the absolute values of the peripheral level differences, evaluates the sum of the absolute values of the peripheral level differences, and determines the inverse quantization table. Output to 047.
- the predictor 048 calculates the predicted value X of the restoration target pixel.
- the quantization number is input to the inverse quantization table determined by the quantization table switching module 047, and the prediction error quantization value is calculated.
- the peripheral pixel B (041) is substituted as the peripheral pixel E (042) of the next restoration target pixel.
- the peripheral pixel C (040) is substituted as the peripheral pixel B (041) of the next restoration target pixel.
- S031 The peripheral pixel D (039) is substituted as the peripheral pixel C (040) of the next restoration target pixel.
- 5032 Obtain peripheral pixel D (039) of the next restoration target pixel from buffer 038 for one line for prediction.
- FIG. 16 is a functional block diagram showing a system configuration of the image data restoration apparatus according to the embodiment of the present invention.
- the image data restoration apparatus inputs the compressed data 202, which is the output of the above-described image data compression apparatus, to the restoration processing unit 210, and the restoration processing unit 210 first inputs the compressed data 202.
- the image data 202 is input to the inverse encoding processing unit 211 corresponding to the inverse encoding unit 032 in FIG. 11, and the quantization number is obtained from the inverse encoding processing unit 211, and this quantization number is converted into the inverse quantizing unit. This is input to the inverse quantization processing unit 212 corresponding to 033.
- the continuity evaluation unit 213 corresponding to the continuity evaluation result feedback module 043 in FIG. 11 in the evaluation unit 212 is such that the level difference between the two specific peripheral pixels related to the restoration target pixel continues. V is evaluated, and if the same level difference continues for a predetermined number of times, a fine control signal for using the inverse quantization table 218 is generated. Also, the sum of absolute values of the level differences of two neighboring pixels including the neighboring pixels of the pixel for which the neighboring evaluation unit 214 corresponding to the neighboring level difference detection module 045 of the evaluating unit 212 calculates the predicted value of the restoration target pixel.
- level difference table 215 that manages the level difference and the information of the quantization table to be used in association with each other is referred to, and the sum of absolute values of the peripheral level differences is evaluated to determine which level is dequantized.
- Control signals indicating whether to use the table 218 are input to the switching processing unit 216, respectively.
- the switching processing unit 216 corresponding to the quantization table switching module 047 instructs the inverse quantization processing unit 217 to switch the inverse quantization table 218 based on the input control signal, and the inverse quantization processing unit 217. Then, the inverse quantization table 218 is switched, and the prediction error quantization value is calculated using the switched quantization table 218 and input to the prediction processing unit 219.
- the prediction processing unit 219 obtains a restored pixel value by adding the predicted value of the restoration target pixel predicted by the predictor and the prediction error quantization value calculated using the switched quantization table 218, and accumulates this. it is intended to obtain a restored (image) data 2 04.
- the processing in the image data restoration apparatus shown in FIG. 16 described above can be realized on a computer.
- various hardware such as various registers, arithmetic units including ALU, RAM, ROM, and I / O are used as computer hardware resources.
- the above-mentioned modules, predictors, inverse quantizers including inverse quantization tables, inverse encoders, etc. are integrated into a single-chip LSI, and various applications that handle two types of image data with different characteristics as described above as image data restoration devices. Is available.
- the image data compression device and the image data restoration device of the present invention can be mounted on a desired application device as a one-chip LSI.
- a desired application device for example, an in-vehicle video data transmission system including a car navigation system and various amusements (games, animations, etc.) ) Possible application to equipment.
- a display is also installed on the rear side. It is desirable to be able to appreciate images different from those on the front side. Therefore, the image data processed by the video data processing device installed on the front side is compressed and transmitted to the rear side, and the rear side restores it and displays the image data. When relaying image data, compression and decompression are repeated at each relay point.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
Abstract
Description
Claims
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2006800546175A CN101444103B (zh) | 2006-05-17 | 2006-08-16 | 图像压缩装置、压缩方法、以及图像恢复装置、恢复方法 |
| JP2008515425A JP4709900B2 (ja) | 2006-05-17 | 2006-08-16 | 画像圧縮装置、圧縮方法及びプログラム並びに画像復元装置、復元方法及びプログラム |
| EP20060796466 EP2034741B1 (en) | 2006-05-17 | 2006-08-16 | Image compression device, compression method, program, and image decompression device, decompression method, and program |
| US12/268,057 US8045814B2 (en) | 2006-05-17 | 2008-11-10 | Image compression device, compressing method, storage medium, image decompression device, decompressing method, and storage medium |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2006138071 | 2006-05-17 | ||
| JP2006-138071 | 2006-05-17 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US12/268,057 Continuation US8045814B2 (en) | 2006-05-17 | 2008-11-10 | Image compression device, compressing method, storage medium, image decompression device, decompressing method, and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2007132539A1 true WO2007132539A1 (ja) | 2007-11-22 |
Family
ID=38693642
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2006/316107 Ceased WO2007132539A1 (ja) | 2006-05-17 | 2006-08-16 | 画像圧縮装置、圧縮方法及びプログラム並びに画像復元装置、復元方法及びプログラム |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US8045814B2 (ja) |
| EP (1) | EP2034741B1 (ja) |
| JP (1) | JP4709900B2 (ja) |
| KR (1) | KR100980354B1 (ja) |
| CN (1) | CN101444103B (ja) |
| WO (1) | WO2007132539A1 (ja) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012029208A1 (ja) * | 2010-09-03 | 2012-03-08 | パナソニック株式会社 | 画像符号化方法、画像復号化方法、画像符号化装置及び画像復号化装置 |
| EP2273776A4 (en) * | 2008-03-31 | 2014-07-16 | Fujitsu Ltd | Image data compression device, decompression device, compression method, decompression method and program |
| JP2018074466A (ja) * | 2016-11-01 | 2018-05-10 | Nttエレクトロニクス株式会社 | 動画像符号化装置および動画像符号化方法 |
| CN114339242A (zh) * | 2021-12-28 | 2022-04-12 | 威创集团股份有限公司 | 一种图像压缩的方法及装置 |
Families Citing this family (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8406314B2 (en) * | 2007-11-28 | 2013-03-26 | Sharp Laboratories Of America, Inc. | Two-dimensional DPCM with PCM escape mode |
| RU2510150C2 (ru) * | 2009-04-03 | 2014-03-20 | Ай-Сес (Инновэйтив Компрешн Инджиниринг Солюшнз) | Способ обработки цифрового файла, в частности, типа изображения, видео и/или аудио |
| US8238680B2 (en) * | 2010-06-25 | 2012-08-07 | Altek Corporation | Image compression method with variable quantization parameters and variable coding parameters |
| TWI405468B (zh) * | 2010-06-25 | 2013-08-11 | Altek Corp | Image Compression Method with Variable Quantization Parameters and Variable Coding Parameters |
| JP5900024B2 (ja) * | 2012-03-02 | 2016-04-06 | 沖電気工業株式会社 | 動画像符号化装置及びプログラム、動画像復号装置及びプログラム、並びに、動画像配信システム |
| JPWO2013154028A1 (ja) * | 2012-04-13 | 2015-12-17 | ソニー株式会社 | 画像処理装置および方法 |
| US9813711B2 (en) | 2012-10-03 | 2017-11-07 | Avago Technologies General Ip (Singapore) Pte. Ltd. | Hybrid transform-based compression |
| US10812829B2 (en) | 2012-10-03 | 2020-10-20 | Avago Technologies International Sales Pte. Limited | 2D block image encoding |
| US9805442B2 (en) | 2012-10-03 | 2017-10-31 | Avago Technologies General Ip (Singapore) Pte. Ltd. | Fine-grained bit-rate control |
| US9978156B2 (en) | 2012-10-03 | 2018-05-22 | Avago Technologies General Ip (Singapore) Pte. Ltd. | High-throughput image and video compression |
| US9883180B2 (en) * | 2012-10-03 | 2018-01-30 | Avago Technologies General Ip (Singapore) Pte. Ltd. | Bounded rate near-lossless and lossless image compression |
| US9363517B2 (en) | 2013-02-28 | 2016-06-07 | Broadcom Corporation | Indexed color history in image coding |
| CN103533351B (zh) * | 2013-09-26 | 2016-08-17 | 西安空间无线电技术研究所 | 一种多量化表的图像压缩方法 |
| GB2533095A (en) * | 2014-12-08 | 2016-06-15 | Cryptomathic Ltd | System and method |
| KR101790773B1 (ko) | 2015-09-10 | 2017-10-26 | 임재각 | 당구 게임용 보조 시스템 |
| KR102601732B1 (ko) | 2016-05-31 | 2023-11-14 | 삼성디스플레이 주식회사 | 영상 부호화 방법 및 영상 복호화 방법 |
| DE102018110383B4 (de) | 2018-04-30 | 2025-11-13 | Basler Ag | Verfahren und Vorrichtung zum Kodieren von Bilddaten |
| EP3928434A1 (en) * | 2019-03-22 | 2021-12-29 | Siemens Corporation | Embedded deep compression for time-series data |
| KR20200129319A (ko) | 2019-05-08 | 2020-11-18 | 주식회사 라이또에프앤씨 | 코칭 경로를 제공하는 당구장 운영시스템 |
| KR20200129321A (ko) | 2019-05-08 | 2020-11-18 | 주식회사 라이또에프앤씨 | 큐대 조준 방향에 따른 당구공 예상 경로를 제공하는 당구장 운영시스템 |
| KR20200129955A (ko) | 2019-05-10 | 2020-11-18 | 주식회사 라이또에프앤씨 | 당구장 점수 자동 계수 및 표시 시스템 |
| GB2593522B (en) * | 2020-03-26 | 2023-02-22 | Imagination Tech Ltd | Image data decompression |
| GB2593523B (en) | 2020-03-26 | 2023-06-14 | Imagination Tech Ltd | Image data compression |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH06350992A (ja) | 1993-06-08 | 1994-12-22 | Sony Corp | データ圧縮回路 |
| JPH10126777A (ja) | 1996-10-21 | 1998-05-15 | Fuji Photo Film Co Ltd | 画像圧縮方法および装置 |
| JP2000059782A (ja) * | 1998-07-21 | 2000-02-25 | Hewlett Packard Co <Hp> | 空間領域デジタル画像の圧縮方法 |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS56136093A (en) * | 1980-03-26 | 1981-10-23 | Fuji Photo Film Co Ltd | Adaptive quantizer |
| DE3331426A1 (de) * | 1983-08-31 | 1985-03-14 | Siemens AG, 1000 Berlin und 8000 München | Anordnung zur zweidimensionalen dpcm-codierung |
| US4613948A (en) * | 1984-06-01 | 1986-09-23 | Bell Communications Research, Inc. | Conditional quantization grey level and color image coding apparatus |
| US4725885A (en) * | 1986-12-22 | 1988-02-16 | International Business Machines Corporation | Adaptive graylevel image compression system |
| US4864397A (en) * | 1987-04-28 | 1989-09-05 | Siemens Aktiengesellschaft | Arrangement for past DPCM coding of video signals according to a 2-D or 3-D coding method |
| WO1991018479A1 (en) * | 1990-05-14 | 1991-11-28 | Eastman Kodak Company | Block adaptive linear predictive coding with adaptive gain and bias |
-
2006
- 2006-08-16 CN CN2006800546175A patent/CN101444103B/zh not_active Expired - Fee Related
- 2006-08-16 EP EP20060796466 patent/EP2034741B1/en not_active Not-in-force
- 2006-08-16 WO PCT/JP2006/316107 patent/WO2007132539A1/ja not_active Ceased
- 2006-08-16 KR KR20087028909A patent/KR100980354B1/ko not_active Expired - Fee Related
- 2006-08-16 JP JP2008515425A patent/JP4709900B2/ja not_active Expired - Fee Related
-
2008
- 2008-11-10 US US12/268,057 patent/US8045814B2/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH06350992A (ja) | 1993-06-08 | 1994-12-22 | Sony Corp | データ圧縮回路 |
| JPH10126777A (ja) | 1996-10-21 | 1998-05-15 | Fuji Photo Film Co Ltd | 画像圧縮方法および装置 |
| JP2000059782A (ja) * | 1998-07-21 | 2000-02-25 | Hewlett Packard Co <Hp> | 空間領域デジタル画像の圧縮方法 |
Non-Patent Citations (2)
| Title |
|---|
| MUSMANN H.G. ET AL.: "Advances in Picture Coding", PROCEEDINGS OF THE IEEE, vol. 73, no. 4, April 1985 (1985-04-01), pages 523 - 548, XP002057834 * |
| See also references of EP2034741A4 |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2273776A4 (en) * | 2008-03-31 | 2014-07-16 | Fujitsu Ltd | Image data compression device, decompression device, compression method, decompression method and program |
| WO2012029208A1 (ja) * | 2010-09-03 | 2012-03-08 | パナソニック株式会社 | 画像符号化方法、画像復号化方法、画像符号化装置及び画像復号化装置 |
| JP2012054882A (ja) * | 2010-09-03 | 2012-03-15 | Panasonic Corp | 画像符号化方法、画像復号化方法、画像符号化装置及び画像復号化装置 |
| JP2018074466A (ja) * | 2016-11-01 | 2018-05-10 | Nttエレクトロニクス株式会社 | 動画像符号化装置および動画像符号化方法 |
| CN114339242A (zh) * | 2021-12-28 | 2022-04-12 | 威创集团股份有限公司 | 一种图像压缩的方法及装置 |
Also Published As
| Publication number | Publication date |
|---|---|
| EP2034741A4 (en) | 2010-12-29 |
| US8045814B2 (en) | 2011-10-25 |
| KR20090008415A (ko) | 2009-01-21 |
| EP2034741A1 (en) | 2009-03-11 |
| US20090052790A1 (en) | 2009-02-26 |
| KR100980354B1 (ko) | 2010-09-06 |
| CN101444103B (zh) | 2013-04-17 |
| JPWO2007132539A1 (ja) | 2009-09-17 |
| CN101444103A (zh) | 2009-05-27 |
| EP2034741B1 (en) | 2013-08-14 |
| JP4709900B2 (ja) | 2011-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2007132539A1 (ja) | 画像圧縮装置、圧縮方法及びプログラム並びに画像復元装置、復元方法及びプログラム | |
| CN101390401B (zh) | 借助伪影评价的提高的图像/视频质量 | |
| US9288453B2 (en) | Method and device for lossy encoding data and corresponding device for reconstructing data | |
| CN101237579B (zh) | 上变换解码的帧的帧频的设备和方法 | |
| US8311107B2 (en) | Image data compression device, compressing method, image data decompression device, decompressing method, and recording medium | |
| CN103458241A (zh) | 用于多层视频编码的速率控制方法和设备 | |
| CN101641961A (zh) | 使用运动补偿滤波的图像编码和解码方法及设备 | |
| US20040101052A1 (en) | Fine granularity scalability encoding/decoding apparatus and method | |
| KR100229796B1 (ko) | 열화영상에 대한 보상기능을 갖는 영상 복호화 시스템 | |
| JP5093349B2 (ja) | 画像圧縮装置及び画像復元装置 | |
| JPWO2012160626A1 (ja) | 画像圧縮装置、画像復元装置、及びプログラム | |
| US20210409744A1 (en) | Image decoding device, image decoding method, and program | |
| JP4485996B2 (ja) | 画像符号化装置及び画像符号化プログラム | |
| JPH11177981A (ja) | 適応的ブロック現象除去機能を有する復号化器及び適応的ブロック現象除去復号化方法 | |
| JP4004597B2 (ja) | 映像信号のエラー隠ぺい装置 | |
| KR0129573B1 (ko) | 디지탈 화상 복호화를 위한 디씨(dc) 계수의 오차보상 방법 | |
| KR20060127159A (ko) | 압축된 디지털 비디오에서 mpeg 손상들의 글로벌표시를 위한 시스템 및 방법 | |
| JPH0358672A (ja) | 画像信号符号化方式及び装置 | |
| GB2329543A (en) | Image comparison in transform-quantised space | |
| JP2002118850A (ja) | 動画像符号化方法、動画像復号化方法、動画像符号化装置および動画像復号化装置 | |
| JP2011120128A (ja) | 画像符号化装置及び画像符号化方法 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 06796466 Country of ref document: EP Kind code of ref document: A1 |
|
| DPE1 | Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101) | ||
| ENP | Entry into the national phase |
Ref document number: 2008515425 Country of ref document: JP Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 200680054617.5 Country of ref document: CN |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 1020087028909 Country of ref document: KR |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2006796466 Country of ref document: EP |