CN1059771C - Digital image decoder and decoding method - Google Patents
Digital image decoder and decoding method Download PDFInfo
- Publication number
- CN1059771C CN1059771C CN95101301A CN95101301A CN1059771C CN 1059771 C CN1059771 C CN 1059771C CN 95101301 A CN95101301 A CN 95101301A CN 95101301 A CN95101301 A CN 95101301A CN 1059771 C CN1059771 C CN 1059771C
- Authority
- CN
- China
- Prior art keywords
- data
- pixel data
- sampling
- sub
- data blocks
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The present invention relates to a device and a method for decoding a digital image. The present invention has the advantages of storage space requirement reduction and storage quantity reduction during high resolution static image decoding. The device is provided with a decompression unit, an image frame storing area and a display unit, wherein the decompression unit is used for reducing compressed image data into reduced pixel square data during fully dynamic image decoding and reducing compressed image data into subsampled pixel square data during high resolution static image decoding; the image frame storing area is used for storing the reduced pixel square data or the subsampled pixel square data, and the reduced pixel square data and the subsampled pixel square data are called by pixel square data as a joint name; the display unit is used for reducing the pixel square data into decompressed image frames.
Description
The invention relates to the digital image decoding system and when decompressing (Decompression), reduce the device and method of memory space requirements, thereby reduce required memory.(Digital Video Decoding System) must use the register mode area of a large amount of memory spaces as the reduction image frame in the digital image decoding system; Other has the buffering area of sub-fraction memory space as packed data; The digital picture decompressing device block diagram that is typical prior art shown in Figure 1; Compressing image data (Compressed VideoData) is input in the input unit 101 to be imported data in the memory cell 100 by input unit 101 again, comprising a packed data buffering area 102 and figure frame memory area (Frame Memory) 104, this packed data buffering area is in order to deposit the compressing image data from input unit 101, for the usefulness of decompression, figure frame memory area 104 then is made for decoding required reference picture and last picture displayed in order to the image frame of depositing reduction; Decompression unit 103 is to do decompression from the data of packed data buffering area, and the data of generation reduction image frame, and these data are deposited in the figure frame memory area 104, the data of figure frame memory area 104 are output as decompress(ion) contract drawing frame via the processing of display unit 105 again; In decode procedure,, control the access of said memory cells 100 respectively, show with the output of reduction image frame with input, the view data decompression operation of carrying out compressing image data by input unit 101, decompression unit 103 and display unit 105.
The size of the memory space that said memory cells 100 is required is relevant with the resolution of employed image compression algorithm and picture, general pattern CD Video CD adopts MPEG (Motion PictureGroup) to be image compression algorithm, wherein full dynamic image is the picture formation by signal source pattern of the input SIF (SourceInput Format) resolution, and the resolution of still image (Still Picture), except the SIF form, other has the high-res still image, so-called high-res is meant that the resolution of picture level and vertical direction respectively is two times of the SIF form, and promptly the number of pixels of a high-res picture is equivalent to the number of pixels sum of four SIF forms.
Above-mentioned MPEG is the international standard of digitlization dynamic image compression, the data structure form of dynamic image compression and the program of decompression thereof in MPEG, have been defined, picture resolution in the SIF form contains 352 * 240 pixels or 352 * 288 pixels, among the MPEG all with Y, G
b, 4: 1: 1 remarked pixels of Cr color, each pixel need have a bit group to represent brightness Y and each 1/4th bit group to represent colourity G
b, Cr, the therefore required maximum storage space that takies of picture of every SIF form, the brightness Y element is 352 * 228=99K bit group, look fills G
b, C
rElement respectively is a 24.75K bit group, and the distribution of its memory is shown among Fig. 2 201,202,203, and it represents Y element, G respectively
bThe storage space of element and Cr element.
Figure 3 shows that the structure of view data in mpeg system, one section consecutive image (VideoSequence) 301 comprises many image frame groups (Group of Pictures) 302, one image frame group comprises many image frames 303 again, and each picture is divided into many slicings (Slice) 304, be divided into many big data blocks (Macrobolck) 305 in the slicing 304 again, and being divided into many data blocks 306 in each big data block 305 again, each represents the Y element or the G of-8 * 8 pixels each data block
bElement or C
rElement.In the display order of picture among the MPEG shown in Fig. 4 A, the image data stream preface (Video Stream Order) of its compression is but as among Fig. 4 B, mpeg picture is needed to be reduced via the mpeg decode system to Fig. 4 A by Fig. 4 B, reduce other picture by in memory, storing reference picture, wherein I represents the original painting picture, do not need with reference to any picture promptly reducible, P represents predictive picture, must can reduce with reference to previous I picture or P picture, and B is a bi-directional predicted picture, must can reduce with reference to two I in front and back or P picture, shown in Fig. 4 C.I among Fig. 4, P are the reference picture that is stored in after the reduction in the memory, and B be utilize before and after the picture that reduced of two reference pictures.Thereby, when the decompression systems of Fig. 1 carries out full moving image decoding, must keep the picture of two oneself reduction as the reference picture that decompresses in the figure frame memory area 104, also need there be a zone to deposit the picture that is reducing simultaneously in the figure frame memory area 104, therefore figure frame memory area 104 need have the memory space that stores three pictures at least, it is equivalent to 445.5k bit group, as shown in Figure 2, with 445.5k bit group, the general usefulness of 4,000,000 (4M) bit dynamic random access memory DRAM of an industrial standard of all using as memory cell 100,, except that the need that satisfy figure frame memory area 104, still there is the space of 66.5k bit group can be made for packed data buffering area 102 and uses with the DRAM of 4M bit; When the decompression systems of Fig. 1 carries out the decoding of high-res still image, because of still image all utilizes original painting coding (Intra-Coding), so do not need in the figure frame memory area 104 as shown in Figure 4 and need reduce other picture in bi-directional predicted mode again as the reference picture that decompresses to go back original picture, but the pixel quantity that is equivalent to 4 times of SIF pictures because of the high-res picture, that is need account for the memory space of four times of SIF pictures, its required memory space as shown in Figure 5, FY501 is that the shared memory space of luminance elements is a 396k bit group, FC
b502, FC
r503 is that the shared memory space of colourity element respectively is 99k bit group, FC
r503, FC
b502, FY501 is that shared altogether storage space size is 594k bit group, this demand has surpassed the capacity of a 4M bit DRAM, thereby must increase memory to deal with when the high-res still image is decoded to the demand of memory cell 100, cause the system hardware cost to improve relatively.
In order to solve the above-mentioned problems in the prior art, main purpose of the present invention is, under the situation that does not increase memory cell, no matter make the digital image decoding system when carrying out the decoding of full moving image decoding or high-res still image, all only need be with the DRAM of one four megabit unit as memory cell, reaching the system works demand, and reduce the cost of system hardware.
For realizing above goal of the invention, the invention provides a kind of digital image decoding device and method, it is with compressing image data decoding and output decompress(ion) contract drawing frame, and this device comprises the restoring data block buffer at least, in order to deposit the reduction pixel data blocks of data that is reduced by compressing image data; The sub-sampling device, tool one start-up control signal, when the digital image decoding device carried out high-res still image decoded state, this start-up control signal was in starting state, and will reduce the pixel data blocks of data and do horizontal sub-sampling and handle, and output sub-sampling pixel data blocks of data; And when carrying out full moving image decoding state in the digital image decoding system, this start-up control signal is in prohibits moving state, and will reduce the straight-through sub-sampling device of pixel data blocks of data and not do the sub-sampling processing; The figure frame memory area is in order to storage and reduction pixel data blocks of data or sub-sampling pixel data blocks of data; This reduction pixel data blocks of data and this sub-sampling pixel data blocks of data general designation pixel data blocks of data; The display buffer device transfers the scan line pixel data in order to the pixel data blocks of data that the figure frame memory area is exported; Interpolation filtering device is in order to make interpolation and Filtering Processing with the scan line pixel data, with output interpolation scan-line data; And form decompress(ion) contract drawing frame by the interpolation scan-line data.
The invention provides the method that a kind of digital image decoding device reduces memory space requirements, comprise the following step: (1) is reduced to reduction pixel data blocks of data with compressing image data: (2) will reduce the pixel data blocks of data and carry out sub-sampling, to obtain sub-sampling pixel data blocks of data; (3) transfer sub-sampling pixel data blocks of data to the scan line pixel data; (4) the scan line pixel data are done interpolation and filtering to obtain the interpolation scan-line data; (5) form decompress(ion) contract drawing frame by the interpolation scan-line data.
For further specifying purpose of the present invention, feature and advantage, be described below in conjunction with most preferred embodiment:
Brief Description Of Drawings:
Fig. 1 is a digital image decoding device block diagram in the prior art.
Fig. 2 is the memory configurations figure of figure frame memory area when carrying out full moving image decoding in the digital image decoding device.
Fig. 3 is the structure of view data in the mpeg system.
Fig. 4 is the decoding of image frame in the mpeg system and shows graph of a relation.
Fig. 5 is the memory configurations figure of figure frame memory area when carrying out the decoding of high-res still image in the digital image decoding device.
Fig. 6 is a digital image decoding device block diagram of the present invention.
Fig. 7 is the memory configurations figure that digital image decoding device of the present invention carries out high-res still image when decoding figure frame memory area.
Fig. 8 is the schematic diagram that reduction pixel data blocks of data is carried out horizontal sub-sampling.
Fig. 9 is the schematic diagram that the scan line pixel data are carried out horizontal interpolation filtering among the present invention.
Figure 10 is a digital image decoding method flow diagram among the present invention.
Fig. 6 is the block diagram of apparatus of the present invention, in the present invention when the digital image decoding device carries out full moving image decoding its action with Fig. 1 described in there is no different, in the compressing image data input unit 601, by input unit 601 data are exported in the memory cell 600 again, wherein, comprise that a packed data buffering area 602 is in order to deposit the compressing image data of using from input unit 601, and figure frame memory area 604 is made for the required reference picture of decoding and the usefulness of display frame in order to the image frame of depositing reduction, decompression unit 603 will be done decompression from the data of packed data buffering area 602, and generation reduction image frame data, and exporting these data to figure frame memory area 604, the data of figure frame memory area 604 are delivered to display unit 605 again and are handled and output decompress(ion) contract drawing frame.The storage space of figure frame memory area 604 distributes as shown in Figure 2, it utilizes reference picture to go back original picture also as shown in Figure 4, characteristics of the present invention are under the memory space requirements of a 4M DRAM, when carrying out the decoding of high-res still image, in decompression unit 603, use the structure and the method for sub-sampling (Subsampling), make reduction image frame data carry out sub-sampling in the horizontal direction in the preceding elder generation of storage map frame memory 604, make 604 need of figure frame memory area deposit pixel as half of 594k bit group in Fig. 5, the distribution of its memory space in figure frame memory area 604 as shown in Figure 7, FY701 is that the brightness Y element is handled the required memory space in back via sub-sampling when the high-res still image is decoded, be 198k bit group, FC
b702 is chrominance C
bElement is handled the required memory space in back via sub-sampling when the high-res still image is decoded, be 49.5k bit group, FC
r703 is chrominance C
rElement is handled the required memory space in back via sub-sampling when the high-res still image is decoded, be 49.5k bit group, and to be 297k bit group be half of 594k bit group among Fig. 5 to the shared altogether storage space among Fig. 7.
In the decoding of high-res still image, compressing image data is input in the input unit 601, and via input unit 601 output and compressing image data is deposited in the packed data buffering area 602, by packed data buffering area 602 compressing image data is passed to decompression unit 603 again, use a restoring data block buffer (Reconstruction Block Buffer) 607 in the decompression unit 603, in order to deposit the pixel data blocks of each reduction, as shown in Figure 8, this reduction pixel data blocks 801 is made up of 8 * 8 pixels, then again by primary sample circuit 608 executive level sub-samplings, then become sub-sampling pixel data blocks 802 and contain 4 * 8 pixels, reduce the purpose of half and reach pixel; Then the data of sub-sampling pixel data blocks just write in the figure frame memory area 604, and so each reduction pixel data blocks is all reduced a half-pix sampling point in the horizontal direction, and therefore the memory space of required figure frame memory area 604 reduces half.Sub-sampling circuit 608 has one to start (Enable) control input signals EN606 in Fig. 6, this signal is directly controlled by the gunzip of decompression unit 603, when carrying out the picture decoding of general SIF form, this start-up control signal EN 606 can be prohibited moving (disable), sub-sampling circuit 608 is not had an effect, thereby all pixel datas will be directly be written among the figure frame memory area 604 when carrying out the high-res still image when decoding by (go-through) sub-sampling circuit 608 in the restoring data block buffer 607,606 of start-up control signal EN are activated, sub-sampling circuit 608 will be carried out the function of sub-sampling at this moment, do not use the filtering function as for sub-sampling algorithm (Algorithm) because of the complexity event of considering circuit, and when level is taken a sample, directly keep a bit in the per in the horizontal direction two adjacent pixel sampling points, remove (drop) a bit, the sampling point of " X " is removed (drop) and falls as shown in Figure 8, and the sampling point of " 0 " then is retained.
In the display unit 605 of Fig. 6, then use a display buffer 609 depositing pixel data, and use an interpolation and filter circuit (Interpolation ﹠amp from figure frame memory area 604; Filtering Circuit) 610 the scan line pixel data of institute's desire output in the display buffer 609 are done interpolative operation and filtering, with reduction original coding image resolution in the horizontal direction, the situation of reducing by interpolation filtering circuit 610 that is shown in Figure 9 through the data of horizontal sub-sampling, the scan line pixel data of being exported by display buffer 609 901 are done interpolative operation and filtering via interpolation filtering circuit 610 and are obtained interpolation scan-line data 902, the picture that is made of interpolation scan-line data 902 is called decompress(ion) contract drawing frame (Decompressed Frames), by display unit 605 outputs.
In the display unit 605, when output shows, show that because of general the demand of resolution is the specification of Digital Television or VGA, thereby no matter the high-res still image of the image of SIF form or the horizontal sub-sampling of process, all need carry out the processing of interpolation and filtering, to meet relatively abundanter resolution demand, so interpolation and filter circuit 610 do not need to start or prohibit because of the different images resolution function of moving (disable) its interpolation filtering.
The pixel of interpolation is the mode that 2 linearities of employing level interleave filtering in the present invention, that is:
O×O h=(A+B) //2
AhB
H is that the pixel of interpolation is that mean value by adjacent two pixel A in front and back, B gets, " // " is expressed as mean value (rounding), get after promptly average and round up, the advantage of this kind practice is simple, can make the circuit complexity of display buffer 609 and interpolation and filter circuit 610 itself minimum, and not lose the quality of high-res still image.
Figure 10 shows that the method flow diagram of digital image decoding system among the present invention, by this flow chart can with the high-res still image also the required memory space of original picture know skill and reduce half, make it be fit to the required specified vol memory of full moving image decoding:
Step 1001 is for being reduced to compressing image data reduction pixel data blocks of data.
Step 1002 takes a decision as to whether the very high resolution image decoding; As being very high resolution image decoding execution in step 1003, as not, then execution in step 1004.
Step 1003 will be reduced the pixel data blocks of data and be carried out sub-sampling, to obtain sub-sampling pixel data blocks of data.
Step 1004 deposits sub-sampling pixel data blocks of data in the figure frame memory in, and the pixel data blocks of data that maybe will reduce deposits in the figure frame memory.
Step 1005 deposits sub-sampling pixel data blocks of data in the figure frame memory or reduction pixel data blocks of data in the display-memory in, and by display-memory output scanning line pixel data.
Step 1006 is done interpolation and filtering with the scan line pixel data, to obtain the interpolation scan-line data.
Step 1007 forms decompress(ion) contract drawing frame by the interpolation scan-line data.
In describing embodiments of the invention and functional-block diagram for ease of allowing the personage who is familiar with this technical field can understand the present invention and make specific term, but design of the present invention more comprises the impartial technology of the purpose of finishing this particular term, all equivalences of being done according to the present invention's design are modified, and all should be encompassed in the claim scope of the present invention.
Claims (6)
1, a kind of digital image decoding device is with compressing image data decoding and output decompress(ion) contract drawing frame, it is characterized in that this device comprises the restoring data block buffer at least, in order to deposit the reduction pixel data blocks of data that is reduced by compressing image data; The sub-sampling device, tool one start-up control signal, when the digital image decoding device carried out high-res still image decoded state, this start-up control signal was in starting state, and will reduce the pixel data blocks of data and do horizontal sub-sampling and handle, and output sub-sampling pixel data blocks of data; And when carrying out full moving image decoding state in the digital image decoding system, this start-up control signal is in prohibits moving state, and will reduce the straight-through sub-sampling device of pixel data blocks of data and not do the sub-sampling processing; The figure frame memory area is in order to storage and reduction pixel data blocks of data or sub-sampling pixel data blocks of data; This reduction pixel data blocks of data and this sub-sampling pixel data blocks of data general designation pixel data blocks of data; The display buffer device transfers the scan line pixel data in order to the pixel data blocks of data that the figure frame memory area is exported; Interpolation filtering device is in order to make interpolation and Filtering Processing with the scan line pixel data, with output interpolation scan-line data; And form decompress(ion) contract drawing frame by the interpolation scan-line data.
2, digital image decoding device according to claim 1 is characterized in that having a data buffer zone, in order to deposit compressing image data, reaches packed data buffering area and described figure frame memory area and synthesizes a memory cell.
3, digital image decoding device according to claim 1, the pixel data blocks of data that it is characterized in that reducing is formed by 8 * 8 pixels.
4, digital image decoding device according to claim 1, the pixel data blocks of data that it is characterized in that reducing is formed by 4 * 8 pixels.
5, a kind of digital image decoding method is characterized in that this method comprises the following step at least:
(1) compressing image data is reduced to reduction pixel data blocks of data;
(2) will reduce the pixel data blocks of data carries out sub-sampling, to obtain sub-sampling pixel data blocks of data;
(3) transfer sub-sampling pixel data blocks of data to the scan line pixel data;
(4) the scan line pixel data are done interpolation and filtering to obtain the interpolation scan-line data;
(5) form decompress(ion) contract drawing frame by the interpolation scan-line data.
6, digital image decoding method according to claim 5, the sub-sampling that it is characterized in that described step (2) is horizontal sub-sampling.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN95101301A CN1059771C (en) | 1995-01-12 | 1995-01-12 | Digital image decoder and decoding method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN95101301A CN1059771C (en) | 1995-01-12 | 1995-01-12 | Digital image decoder and decoding method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1126927A CN1126927A (en) | 1996-07-17 |
CN1059771C true CN1059771C (en) | 2000-12-20 |
Family
ID=5073894
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN95101301A Expired - Fee Related CN1059771C (en) | 1995-01-12 | 1995-01-12 | Digital image decoder and decoding method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN1059771C (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040010329A1 (en) * | 2002-07-09 | 2004-01-15 | Silicon Integrated Systems Corp. | Method for reducing buffer requirements in a digital audio decoder |
KR100498362B1 (en) * | 2003-08-07 | 2005-07-01 | 엘지전자 주식회사 | Method for post-processing decoded image of mobile phone |
KR100498363B1 (en) * | 2003-08-07 | 2005-07-01 | 엘지전자 주식회사 | Method for post-processing decoded image of mobile phone |
CN100356781C (en) * | 2004-07-07 | 2007-12-19 | 华晶科技股份有限公司 | Method and apparatus for sampling dynamic image |
US8340195B2 (en) * | 2010-04-09 | 2012-12-25 | Newport Media, Inc. | Buffer size reduction for wireless analog TV receivers |
JP5972687B2 (en) * | 2012-07-02 | 2016-08-17 | 株式会社Nttドコモ | Moving picture predictive coding apparatus, moving picture predictive coding method, moving picture predictive coding program, moving picture predictive decoding apparatus, moving picture predictive decoding method, and moving picture predictive decoding program |
CN108235024B (en) * | 2018-02-05 | 2020-02-07 | 杭州视为科技有限公司 | Method and device for compressing image |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1088041A (en) * | 1992-10-28 | 1994-06-15 | 菲利浦电子有限公司 | Device and corresponding decoding device to the encoding digital signals of representing images |
-
1995
- 1995-01-12 CN CN95101301A patent/CN1059771C/en not_active Expired - Fee Related
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1088041A (en) * | 1992-10-28 | 1994-06-15 | 菲利浦电子有限公司 | Device and corresponding decoding device to the encoding digital signals of representing images |
Also Published As
Publication number | Publication date |
---|---|
CN1126927A (en) | 1996-07-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5835636A (en) | Method and apparatus for reducing the memory required for decoding bidirectionally predictive-coded frames during pull-down | |
US5859651A (en) | Method and apparatus for block data transfer to reduce on-chip storage for interpolative video resizing | |
JP3943129B2 (en) | Using memory to decode and display video with 3: 2 pulldown | |
US5581310A (en) | Architecture for a high definition video frame memory and an accompanying data organization for use therewith and efficient access therefrom | |
US6088047A (en) | Motion compensated digital video decoding with buffered picture storage memory map | |
US6088391A (en) | Method and apparatus for segmenting memory to reduce the memory required for bidirectionally predictive-coded frames | |
US5912710A (en) | System and method for controlling a display of graphics data pixels on a video monitor having a different display aspect ratio than the pixel aspect ratio | |
DE19756210C2 (en) | Method and apparatus for decoding an encoded MPEG video data stream | |
US6046773A (en) | Apparatus and method for decoding video images | |
US5742272A (en) | Accelerated full screen video playback | |
EP0264726B1 (en) | Picture transformation memory | |
CN101212674B (en) | Image address mapping method in memory | |
US6215822B1 (en) | Motion compensated digital video decoding and buffer memory addressing therefor | |
JPH11331751A (en) | Tiling of picture memory map for minimizing memory bandwidth of compression and decompression of data sequence | |
US5977947A (en) | Method and apparatus for resizing block ordered video image frames with reduced on-chip cache | |
JPH08237662A (en) | Picture element interpolation filter for video decompressionprocessor | |
US5541658A (en) | Image coding-decoding apparatus with efficient memory access | |
EP0729276B1 (en) | Memory control system and video decoder using the same | |
CN1164078C (en) | Character frame method for 16 : 9 aspect ratio transformation of MPEG image | |
US5812144A (en) | System for performing real-time video resizing in a data processing system having multimedia capability | |
KR100359039B1 (en) | Image processing apparatus and image memory mapping method | |
CN1059771C (en) | Digital image decoder and decoding method | |
JPH10271529A (en) | Image processor, still image pickup device and image processing method | |
CN1169345C (en) | Method and apparatus for mapping a digital versatile Disk (DVD) image onto high resolution computer display device | |
KR100877164B1 (en) | A method, an apparatus and a video system for edge filtering video macroblocks |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
C19 | Lapse of patent right due to non-payment of the annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |