CN106488240A - Image processing system and image treatment method - Google Patents
Image processing system and image treatment method Download PDFInfo
- Publication number
- CN106488240A CN106488240A CN201510551785.6A CN201510551785A CN106488240A CN 106488240 A CN106488240 A CN 106488240A CN 201510551785 A CN201510551785 A CN 201510551785A CN 106488240 A CN106488240 A CN 106488240A
- Authority
- CN
- China
- Prior art keywords
- image
- block
- image processing
- random access
- group
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09G—ARRANGEMENTS OR CIRCUITS FOR CONTROL OF INDICATING DEVICES USING STATIC MEANS TO PRESENT VARIABLE INFORMATION
- G09G5/00—Control arrangements or circuits for visual indicators common to cathode-ray tube indicators and other visual indicators
- G09G5/36—Control arrangements or circuits for visual indicators common to cathode-ray tube indicators and other visual indicators characterised by the display of a graphic pattern, e.g. using an all-points-addressable [APA] memory
- G09G5/39—Control of the bit-mapped memory
- G09G5/393—Arrangements for updating the contents of the bit-mapped memory
Abstract
The present invention provides an image processing system and image treatment method, comprises image processing module, frame buffering coding module and frame buffer.Each image block comprises multiple first kind encoded block and at least one Equations of The Second Kind encoded block.Image processing module produces the first image processing result according to multiple first kind encoded block of target image block.Frame buffering coding module produces the first frame buffering coding result according to the first image processing result.Frame buffer is provided with one first staging area of tool at least one first random access point and one second staging area of tool at least one second random access point for target image block.First frame buffering coding result is stored into this first staging area；At least one Equations of The Second Kind encoded block of target image block is stored into this second staging area.
Description
Technical field
The present invention is related to image processing technique, and especially related in order to store the frame buffer of image data.
Background technology
Flourishing with all kinds of digital electronic products, current audio-visual multimedia data is with digital lattice mostly
Formula exists.The digital archives of video-audio data are generally quite huge, and the higher person's data volume of resolution is bigger.Citing comes
Say, in the application of visual telephone or DTV etc., these video-audio datas are generally in the way of crossfire via wireless
Network transmission,Due toData volume is very high, may lead to the not enough problem of transmitting bandwidth.Therefore, can adopt in transmission end
With digital audio-video coding techniques will these video-audio datas encode after row transmission again, widely used at present digital audio-video
Coding techniques is, for example, MPEG-4, H.264 etc..
When receiving terminal receive by audio-visual coding after data, make decoder by encode after data decoded back be
The video-audio data that transmission end sends.Many receiving terminals can use double data rate Synchronous Dynamic Random Access Memory
(Double Data Rate Synchronous Dynamic Random Access Memory,DDR SDRAM)
To keep in the video-audio data after reduction Deng storage device as buffer (buffer), for follow-up image processing journey
Sequence uses.
Decoded video-audio data comprises multiple frame of video, is stored in frame buffer, and each frame of video is defined
For having multiple image block, the unit that each image block is accessed as frame of video, for carrying out follow-up image
Frame of video in processing routine, such as Fig. 1 (A) comprises image block (0,0)～image block (M, N), and every
The size of individual image block is fixing.
However, because the data volume of video-audio data is huge, spent frame buffer space is also big；On the other hand,
When carrying out image processing program, if wishing to improve the speed accessing video-audio data from frame buffer, just need to adopt
Frequency range is larger but also memorizer costly.The holistic cost that thus undoubtedly can cause hardware system rises.
Content of the invention
For solving the above problems, the present invention proposes a kind of new video frame buffer and dynamic image processing system.
A specific embodiment according to the present invention is a kind of image processing system, in order to enter for multiple image block
Row one image processing program.Each image block in the plurality of image block comprise multiple first kind encoded block with
At least one Equations of The Second Kind encoded block.This image processing system comprises an image processing module, a frame buffers coding module,
One frame buffer and a frame buffer controller.This image processing module is to the multiple first kind in a target image block
Encoded block imposes image processing, to produce one group of first image processing result.This frame buffering coding module is to this group
First image processing result imposes frame buffering coding, to produce one group of first frame buffering coding result.This frame buffer
It is provided with one first staging area and one second staging area corresponding to this target image block.This first staging area
There is at least one first random access point.This second staging area has at least one second random access point, independent of
This at least one first random access point.This frame buffer controller makes this group first frame buffering coding result be stored into this frame
This first staging area in buffer, and make at least one Equations of The Second Kind encoded block of this target image block be deposited
Enter this second staging area in this frame buffer.
Another specific embodiment according to the present invention is a kind of image treatment method, in order to coordinate one frame buffer pin
Multiple image block are carried out with an image processing program.Each image block in the plurality of image block comprises multiple
First kind encoded block and at least one Equations of The Second Kind encoded block.This frame buffer is provided with corresponding to a target image block
One first staging area and one second staging area.This first staging area has at least one first random access point.
This second staging area has at least one second random access point, independent of this at least one first random access point.First
First, the multiple first kind encoded block in this target image block are subjected to image processing, to produce one group of first shadow
As result.Then, this group the first image processing result is subjected to frame buffering coding, to produce one group of first frame
Buffering coding result.Subsequently, this group first frame buffering coding result is stored into this first working area in this frame buffer
Domain, and at least one Equations of The Second Kind encoded block of this target image block be stored in this frame buffer this second keep in
Region.
Can be further understood by following detailed description and accompanying drawings with regard to the advantages and spirit of the present invention.
Brief description
Fig. 1 (A) is the schematic diagram that a picture is divided into multiple image block.
Fig. 1 (B) is the schematic diagram that an image block is divided into multiple encoded block.
Fig. 2 (A) assumes storage area planning mode in a kind of typical frame buffer.
Fig. 2 (B) assumes a kind of planning mode of storage area in frame buffer of cooperation data compression.
Fig. 3 (A)～Fig. 3 (C) is the functional block diagram according to the image processing system in one embodiment of the invention.
Fig. 4 (A)～Fig. 4 (D) taking eight encoded block of an image block as a example, assumes the frame according to the present invention
Several examples of random access point and the first staging area, the relativeness of the second staging area in buffer.
Fig. 5 is the functional block diagram according to the image processing system in another embodiment of the present invention.
Fig. 6 assumes a kind of schematic diagram that encoded block is divided into multiple prediction block.
Fig. 7 is the flow chart according to the image treatment method in one embodiment of the invention.
Symbol description
300、500：Image processing system 31,51：Frame buffer
32、52：Frame buffer controller 33,53：Image-decoding module
34、54：Image processing module 35,55：Frame buffers coding module
410：First staging area 420：Second staging area
S71～S74：Process step
It should be noted that, the accompanying drawing of the present invention comprises to assume the function block of multiple functional modules associated with each other
Figure.These accompanying drawings not thin portion circuit diagram, and connecting line therein is only in order to represent signal stream.Functional element and
/ or program between multiple interactive relationship be not necessarily intended to through direct electrical connection begin can reach.Additionally, it is first individually
The mode that the function of part is not necessarily intended to as illustrated in schema is distributed, and distributed block is not necessarily intended to distributed
Electronic component is realized.
Specific embodiment
According in one embodiment of the invention, each image block is further defined as multiple encoded block again,
For example image block (0,0) is further defined as the encoded block 0～encoded block 7 in Fig. 1 (B).This enforcement
In example, encoded block is the least unit of accessing video frame.Fig. 2 (A) illustrates an image block
Unit is stored in frame buffer.In case of the data volume of these encoded block respectively 64 bit groups,
Assume storage area planning mode in a kind of frame buffer.In this example, eight volumes of image block (0,0)
Code block 0～7 is (with symbol EB0(0,0)、EB1(0,0)、…、EB7(0,0)Represent, be referred to as [EB0～EB7](0,0)) each
Being stored in a length is in the storage space of 64 bit groups.
It is a kind of image processing system 300 according to one embodiment of the invention, its functional block diagram is illustrated in Fig. 3 (A).
Image processing system 300 comprise one frame buffer 31, a frame buffer controller 32, an image-decoding module 33,
One image processing module 34 and frame buffering coding module 35.Image processing system 300 is in order to for multiple images
Image block (0,0) in block, such as Fig. 1 (A)～image block (M, N), carries out image processing program.
Paragraphs below mainly taking Fig. 1 (A) image block dividing mode as a example to illustrate having in the technical field of the invention
Usually intellectual passes through and subsequently illustrates it is understood that scope of the invention is not limited with above-mentioned block dividing mode.
Framework please also refer to the image block of Fig. 2 (A) and the image processing system 300 of Fig. 3 (A).Entering
During row image processing program, common situation is, because the encoded block 6,7 of image block (0,0) is original
Adjacent with image block (1,0) in picture, for the target image block to be processed, that is, image block
(1,0), when carrying out image processing program in image processing module 34, need to be by image block (0,0) as ginseng
Examine image block.In more detail, it is the data of the encoded block 6,7 with reference to image block (0,0).Assume
Image block (0,0) has been stored into frame buffer 31 a little earlier, and image processing module 34 is for image block (1,0)
When carrying out image processing program, for example, go blocking program, frame buffer controller 32 is just according to image processing module
34 request, from frame buffer 31 by the data read-out of the encoded block 6,7 of image block (0,0) to image
Processing module 34.In one embodiment, code segment block rather than whole image block are only read in order to reach
Purpose, eight storage space depositing the encoded block 0～7 of image block (0,0) are each assigned with one
Random access point (random access point), such as the random access point 0～7 in Fig. 2 (A).Frame buffering control
Device 32 processed is therefore able to the request according to image processing module 34, is directly accessed image area from frame buffer 31
Any one encoded block of block (0,0), starts to read from the encoded block being located at address earlier above without first.
In one embodiment of the invention as shown in Fig. 3 (A), decoded video-audio data is in being stored in frame buffer 31
Before, coding module 35 can be buffered by frame and carry out frame buffering coding, then the video-audio data after frame buffering coding is deposited
Enter frame buffer 31, can save and access required transmitting bandwidth during video-audio data from frame buffer 31.Herein
Alleged frame buffering coding can be a kind of compress mode, and here does not carry out any restriction, compress mode to compress mode
Also should not limit scope of the invention.
Additionally, Fig. 2 (B) assumes storage area planning mode in a kind of frame buffer of cooperation data compression.At this
In individual example, corresponding to data after eight compressions of eight encoded block of image block (0,0) (with symbol [EB0~
EB7]”(0,0)Represent) differ in size.As shown in Fig. 2 (B), data [EB after compression0～EB7]”(0,0)From random access memory
Point 0 starts to be stored in frame buffer by continuous no compartment of terrain.In this example, data EB after compression0”(0,0)
With data EB after compression1”(0,0)Data volume summation be 64 bits, therefore jointly occupy and open from random access point 0
Beginning first row length is the storage space of 64 bit groups.Data EB after compression2”(0,0)Data volume be less than 64
Data EB after bit, the therefore compression of some3”(0,0)(it is denoted as EB3A”(0,0)) with compression after data EB2”(0, 0)Jointly occupy the storage space that secondary series length is 64 bit groups.Similarly, data after the compression of another part
EB3”(0,0)(it is denoted as EB3B”(0,0)) with a part compression after data EB4”(0,0)(it is denoted as EB4A”(0,0)))
Jointly occupy the storage space that the 3rd row length is 64 bit groups.The rest may be inferred, the pressure in the example of Fig. 2 (B)
Data [EB after contracting0～EB7]”(0,0)Start from random access point 0, altogether occupying six row storage space (only makes
With the 6th a part of and not all row).
As known to persond having ordinary knowledge in the technical field of the present invention, can be read out from frame buffer every time
Data volume be definite value, such as 32 bit groups or 64 bit groups.Can be read out from frame buffer every time
Data volume be 32 bit groups (corresponding to half row storage space) in case of, compare Fig. 2 (A) and Fig. 2 (B)
Can be seen that, such as need to be read the data related to encoded block 0～7, the encoded block shown in Fig. 2 (A) needs altogether
It is read the data of 512 bit groups (8 row), and the coding region shown in Fig. 2 (B), after frame buffering coding
The data that 352 bit groups (5.5 row) only from the beginning of random access point 0, need to be read of block.Obviously, adopt
The encoded block through frame buffering coding as shown in Fig. 2 (B) can reach the effect saving volume of transmitted data.So
And, this way of Fig. 2 (B) the disadvantage is that, after only compressing data [EB6～EB7]”(0,0)Need to be read out
When coming, because only that random access point 0 is present in the original position (data after namely compressing of frame buffer
EB0”(0,0)Original position), data [EB upon compression6～EB7]”(0,0)Before being read out, number after compression
According to [EB0～EB5]”(0,0)Also need all to be read out, instead result in reading efficiency low and read frequency range wave
Take.
In another embodiment in accordance with the invention, the encoded block included in image block is divided into two classes.
More particularly, each each self-contained multiple first kind encoded block of image block and at least one Equations of The Second Kind coding region
Block.The mode of these encoded block classification is based on following factor：It is intended to an image block (referred to as target image block)
Part when carrying out image processing, a part for another image block (referred to as with reference to image block) also must be read
For above-mentioned image processing program.Accordingly, a part for above-mentioned target image block is considered first kind encoded block,
An above-mentioned part with reference to image block is considered Equations of The Second Kind encoded block.When the Equations of The Second Kind coding with reference to image block
Block need to be read, and the first kind encoded block of this reference image block is not required under the situation being read, and this second
Class encoded block can be individually read from frame buffer 31.H.264 presented with Fig. 1 (B) with image coding specification
As a example encoded block dividing mode, the encoded block 0～5 in each image block can be considered first kind coding
Block, and encoded block 6,7 can be considered Equations of The Second Kind encoded block.Refer to Fig. 3 (A), the present invention one is implemented
In the image processing system 300 of example, for each image block in processing, frame buffer 31 is provided with correspondence
The first staging area in first kind encoded block and the second staging area corresponding to Equations of The Second Kind encoded block.Should
First staging area has at least one first random access point, and to have at least one second random for this second staging area
Access point.By the setting of different random access points, in the first kind coding region corresponding to same image block
In the case that block is not read, Equations of The Second Kind encoded block still can individually be read from frame buffer 31 that is to say, that
Before reading Equations of The Second Kind encoded block, it is not required to first read first kind encoded block.Deposit at random according to the first of the present invention
Take and a little subsequently will coordinate Fig. 4 (A)～Fig. 4 (D) explanation with the set-up mode example of the second random access point.
As shown in Fig. 3 (A), in image processing system 300, image-decoding module 33 is responsible for input data I
Impose image-decoding, to obtain data (GTG of such as each pixel, chroma, the brightness ... of each encoded block
Deng).Relative position in affiliated image block for each encoded block is known person in advance.Therefore, image solution
Code module 33 can determine whether that each encoded block decoding is first kind encoded block or Equations of The Second Kind encoded block.
Inhomogeneous encoded block can be transferred to different signal paths by the outfan of image-decoding module 33.Clearer and more definite
Ground says, for same image block, image-decoding module 33 provides its first kind encoded block to image
Reason module 34, and through the control of frame buffer controller 32, its Equations of The Second Kind encoded block is stored in frame buffer 31.
Frame buffer controller 32 be responsible for receive modules for frame buffer 31 access request, coordinate frame buffer
The right to use of 31 data access port, and can numbering corresponding access initial address according to each image block
It is supplied to the module filed a request.For example, wish some image area receiving image-decoding module 33
After the Equations of The Second Kind encoded block of block is stored in the request of frame buffer 31, frame buffer controller 32 meeting will be corresponding to this shadow
Storage initial address as the Equations of The Second Kind encoded block of block is supplied to image-decoding module 33, makes this or many
Individual Equations of The Second Kind encoded block is stored into the second staging area corresponding to this image block in frame buffer 31.Connecing
After receiving the first kind encoded block of image-decoding module 33 offer, image processing module 34 can be to these first kind
Encoded block imposes image processing, to produce one group of first image processing result.For example, image processing module
34 image processing being carried out can go blocking program for one.Subsequently, frame buffering coding module 35 is responsible will be right
This group the first image processing result imposes frame buffering coding, to produce one group of first frame buffering coding result, and thoroughly
Cross the assistance of frame buffer controller 32, this group first frame buffering coding result is stored in frame buffer 31 and corresponds to
First staging area of these image block belonging to first kind encoded block.
Additionally, the situation that Fig. 3 (A) processes the image block (1,0) in Fig. 1 (A) with image processing system 300 is
Example, indicates the signal content of each intermodule transmission.Assume that the encoded block 0～5 in each image block is
First kind encoded block, encoded block 6,7 is Equations of The Second Kind encoded block.Also, as discussed previously it is assumed that image
Processing module 34 needs to refer to one when the first kind encoded block to a certain target image block carries out image processing
Equations of The Second Kind encoded block with reference to image block.As shown in Fig. 3 (A), image-decoding module 33 completes image area
After the decoding of block (1,0), by the first kind encoded block [EB of image block (1,0)0～EB5](1,0)Pass to image
Processing module 34, and the Equations of The Second Kind encoded block [EB by image block (1,0)6～EB7](1,0)Pass to frame buffering
Device 31.If image block (0,0) is its reference image block with respect to image block (1,0), and image block
(0,0) two Equations of The Second Kind encoded block [EB6～EB7](0,0)Delay a little earlier frame being stored in by image-decoding module 33
Rush device 31, then image processing module 34 can be filed a request to frame buffer controller 32, with from frame buffer 31
The Equations of The Second Kind encoded block of image block (0,0) is read out in the second staging area corresponding to image block (0,0)
[EB6～EB7](0,0).Then, image processing module 34 and frame buffering coding module 35 be responsible for sequentially produce accordingly right
Should be in one group of first image processing result [EB of image block (1,0)0～EB5]’(1,0)And one group of first frame delays
Rush coding result [EB0～EB5]”(1,0).Subsequently, this group first frame buffering coding result [EB0～EB5]”(1,0)Will be by
It is stored in the first staging area corresponding to image block (1,0) in frame buffer 31.
In an embodiment, image processing module 34 is in the first kind encoded block [EB to image block (1,0)0~
EB5](1,0)When carrying out image processing, also can be to the Equations of The Second Kind encoded block [EB of image block (0,0)6～EB7](0,0)
Carry out image processing, produce one group of second image processing result [EB6～EB7]’(0,0).With image processing module 34
Function be go blocking in case of, image processing module 34 also may in the lump with reference to image block (1,0)
First kind encoded block [EB0～EB5](1,0)And the Equations of The Second Kind encoded block [EB of image block (0,0)6~
EB7](0,0)To change the Equations of The Second Kind encoded block [EB of image block (0,0)6～EB7](0,0).In the case of this, on
State the second image processing result [EB6～EB7]’(0,0)The namely Equations of The Second Kind encoded block after blocking through the past
[EB6～EB7](0,0).As shown in Fig. 3 (B), this group the second image processing result [EB6～EB7]’(0,0)Be stored back to
The second staging area of image block (0,0) is corresponded in frame buffer 31.Or, such as shown in Fig. 3 (C), should
Organize the second image processing result [EB6～EB7]’(0,0)Transferred to frame buffering coding module 35 to carry out frame buffering to be encoded to
One group of second frame buffers coding result [EB6～EB7]”(0,0).Subsequently, frame buffering coding module 35 can buffer to frame
Controller 32 is filed a request, to make the second frame buffering coding result [EB6～EB7]”(0,0)It is subsequently saved back frame buffer 31
In corresponding to image block (0,0) the second staging area.It should be noted that, the frame buffering coding staff of image data
Formula is known to persond having ordinary knowledge in the technical field of the present invention, does not repeat in this.
As it was earlier mentioned, first kind encoded block is before after image processing, being stored into frame buffer 31,
Coding module 35 will be buffered by frame to compress, and Equations of The Second Kind encoded block, after image processing, is stored into frame
Before buffer 31, the compression that also may buffer coding module 35 through frame is processed, such as shown in Fig. 3 (C).
As a example Fig. 4 (A)～Fig. 4 (D) is with eight encoded block 0～7 of image block (1,0), assume frame buffer 31
Middle random access point and several examples of the first staging area, the relativeness of the second staging area.Scheme at this four
In example, the first staging area in order to store first kind encoded block is denoted as 410, compiles in order to store Equations of The Second Kind
Second staging area of code block is denoted as 420.In practice, first kind encoded block and Equations of The Second Kind encoded block
Quantity be not learn person in advance, and the first staging area 410 and the appearance of the second staging area 420 can be set accordingly
Amount.Taking the encoded block dividing mode that high-effect image coding (HEVC) specification and Fig. 1 (B) present as a example, if often
The length and width of one image block are respectively 32 pixel * 16 pixel, the then number before eight encoded block 0～7 are compressed
It is respectively 64 bit groups according to amount size.In the case of using lossless formula (lossless) compress technique, first
Staging area 410 is designed to comprise the storage space that six length is 64 bit groups, and the second working area
Domain 420 is designed to comprise the storage space that two length is 64 bit groups, to guarantee no matter actual compression
Ratio why, all be enough to accommodate an image block behind the first staging area 410 and the second staging area 420 totalling
Partial data.
Referring firstly to Fig. 4 (A).In this example, it is 64 bit groups that the first working area 410 comprises eight length
Storage space.Frame buffers the first framing buffering coding result [EB that coding module 35 provides0～EB5]”(1,0)
Each it is stored in the first staging area 410.Be can be seen that by Fig. 4 (A), this five frames buffer coding result [EB0~
EB5]”(1,0)Correspond respectively to an encoded block, and each by frame buffer controller 32 assign with one first with
Machine access point (1-0,1-1 ..., 1-5).Two Equations of The Second Kind encoded block [EB that image-decoding module 33 provides6~
EB7](1,0)It is then to be stored into the second staging area 420, and each assigned with one second by frame buffer controller 32
Random access point (2-0,2-1).In that case, the modules in image processing system 300 can claim frame
Buffer control unit 32 is assisted, the independent access frame buffer coding result [EB from frame buffer 310～EB5]”(1,0)
And encoded block [EB6～EB7](1,0)Arbitrary pen data in this eight pen data.
Refer to Fig. 4 (B).In this example, frame buffers the first framing buffering coding knot that coding module 35 provides
Really [EB0～EB5]”(1,0)It is stored in the first staging area 410 by continuous no compartment of terrain, and have one first to deposit at random
Take point (1-0).That is, frame buffering coding result EB0”(1,0)Afterwards can back to back storage space meeting quilt
In order to store frame buffering coding result EB1”(1,0), the rest may be inferred.Until certain string storage space is full, frame delays
Rush the next column storage space storage data that controller just can start from the first staging area 410.As Fig. 4 (B)
Shown, the first row storage space in the first staging area 410 just stores frame buffering coding result simultaneously
EB0”(1,0)Buffer coding result EB with frame1”(1,0), and the secondary series storage space in the first staging area 410
Store frame buffering coding result EB simultaneously2”(1,0)Buffer coding result EB with a part of frame3”(1,0)(it is denoted as
EB3A”(1,0)).In this example, the first framing buffering coding result [EB0～EB5]”(1,0)Occupy first temporary
Deposit the first～tetra- row storage space (only using the 4th a part of and not all row) in region 410.The opposing party
Face, two encoded block [EB6～EB7](1,0)The string memorizer being each stored in the second staging area 420 is empty
Interior, and each assigned by frame buffer controller 32 with second random access point (2-0,2-1).This do
Method is advantageous in that, encoded block [EB6～EB7](1,0)Any one of encoded block can be individually read.
Refer to Fig. 4 (C).In this example, frame buffers the frame buffering coding result [EB that coding module 35 provides0~
EB5]”(1,0)It is divided into five parts, be each stored in the first staging area 410, and each by frame buffer controller
32 assign with first random access point (1-0,1-1 ..., 1-5).On the other hand, buffer coding without frame
Two Equations of The Second Kind encoded block [EB6～EB7](1,0)It is then that the second staging area 420 is stored in by continuous no compartment of terrain
Interior, and share one second random access point (2-0).
Refer to Fig. 4 (D).In this example, frame buffers the frame buffering coding result [EB that coding module 35 provides0~
EB5]”(1,0)It is stored in the first staging area 410 by continuous no compartment of terrain, and share one first random access point
(1-0).Similarly, without two Equations of The Second Kind encoded block [EB of frame buffering coding6～EB7](1,0)By continuously continuously
It is stored in the second staging area 420 every ground, and share one second random access point (2-0).In practice, frame delays
Rush whether each encoded block can be buffered coding result through frame buffering coding and each frame by coding module 35
Data volume is recorded, and is stored in frame buffer 31 or in other buffers.In other words, frame buffer 31
In can be datum corresponding to the data volume of each encoded block.Assume that frame buffer controller 32 can delay from frame every time
The data volume rushing device 31 reading is fixed as 32 bit groups (corresponding to half row storage space of in figure).Just scheme
For the example that 4 (D) present, it is intended to when frame buffer 31 reads image block (1,0), frame buffer controller 32
Coding result [EB can be buffered according to frame0～EB5]”(1,0)Respective data volume estimates frame buffering coding result
[EB0～EB5]”(1,0)Occupy the first～tetra- row storage space in the first staging area 410, and use more than one
Half the 4th row storage space.On the other hand, without the encoded block [EB of frame buffering coding6～EB7](1,0)Occupy
First row in second staging area 420 and secondary series storage space.Therefore, frame buffer controller 32 is first
32 bit groups should be read for continuous eight times and (read 256 bit groups altogether from the first random access point 1-0
Data), just frame can be buffered coding result [EB0～EB5]”(1,0)Read.Subsequently, frame buffer controller 32 again from
Second random access point 2-0 starts, and reads 32 bit groups (reading the data of 96 bit groups altogether) for continuous four times,
Just can be by encoded block [EB6～EB7](1,0)Read.On the other hand, read encoded block [EB from frame buffer 316~
EB7](1,0)When, frame buffer controller 32 can start directly from the second random access point 2-0, reads 32 continuous four times
Individual bit group (reading the data of 96 bit groups altogether).
It should be noted that because Equations of The Second Kind encoded block is stored in second different from the first staging area 410
Staging area 420, reads, changes, compresses and restores after compression independent of outside first kind encoded block after compression
Equations of The Second Kind encoded block is feasible.If additionally, no individually reading frame buffering coding result [EB0～EB5]”(1,0)
Middle any frame buffers the demand of coding result, makes frame buffering coding result [EB0～EB5]”(1,0)With same first with
Machine access point can't throw into question, that is, does not need for each frame buffering coding result to be separately stored in different storages
Device space and one random access point of respective assignment.
It should be noted that, the Equations of The Second Kind encoded block being stored in the second staging area 420 is given image processing
Module 34 is a kind of probability of use Equations of The Second Kind encoded block as reference data.These Equations of The Second Kind coding regions
Block also can be used by other modules in image processing system 300.
Fig. 5 is the functional block diagram according to the image processing system in another embodiment of the present invention.Image processing
Main difference is that of image processing system 300 in system 500 and Fig. 3 (A), image-decoding module 53 decodes
The encoded block producing all can be first stored in frame buffer 51.Treat that image processing module 54 needs some codings
During the data of block, just frame buffer controller 52 is filed a request, from frame buffer 51, these data are taken out.
Similarly, the first image processing result that image processing module 54 produces can transfer to frame buffering coding module 55 to compress
Afterwards, restore frame buffer 51.Identical with frame buffer 31, frame buffer 51 is also in each process
Image block is provided with corresponding to the first staging area of first kind encoded block and corresponding to Equations of The Second Kind encoded block
The second staging area.This first staging area has at least one first random access point, and this second staging area
There is at least one second random access point.Thereby, even if in the first kind coding region corresponding to same image block
In the case that block is not read, Equations of The Second Kind encoded block still can individually be read from frame buffer 51.
Persond having ordinary knowledge in the technical field of the present invention is not it is understood that scope of the invention is with above-mentioned example
In number of blocks, data volume size, block dividing mode, the storage space method of salary distribution etc. be limited.
In practice, each encoded block can be as shown in fig. 6, again each be divided into comprising multiple prediction block.
In the example of Fig. 6, an encoded block comprises four prediction block.For example it is assumed that compiling as dynamic image
Each image block 4 of code ultimate unit grows a width of 32 pixel * 16 pixel, the then length and width of each encoded block
Can be 16 pixel * 4 pixel, and the length and width of each prediction block can be 4 pixel * 4 pixel.Prediction block can be set
It is set to a minimal compression/decompression unit block.That is, each prediction block is each in the plurality of prediction block
Compress information, the such as bit length (bit length) after compressing from having corresponding one.
It should be noted that, if the compression that encoded block encodes through frame buffering before being stored in frame buffer 31,51,
Be subsequently removed using before can transfer to one frame buffering decoder module (not illustrating) decompression.In an embodiment, frame
Buffering coding module 35,55 optionally imposes frame to the image processing result of image processing module 34,54 output
Buffering coding, and decide whether an image processing result is compressed according to a compression efficiency threshold value.Citing and
It is assumed that this compression efficiency threshold value is 70%, frame buffering coding module 35 is entering speech for an image processing result
After compression can first be assessed before row compression, whether data volume can be less than 70% of data volume before compression.If it is not, then frame buffers
Coding module 35 does not just impose condensing routine to this image processing result.This way be advantageous in that can save right
The not good image processing result of compression efficiency carries out frame and buffers electric power and time required for the buffering decoding of coding/frame.
It is a kind of image treatment method according to another embodiment of the present invention, many in order to coordinate one frame buffer to be directed to
Individual image block carries out an image processing program, and its flow chart is illustrated in Fig. 7.Each in the plurality of image block
Image block comprises multiple first kind encoded block and at least one Equations of The Second Kind encoded block.This frame buffer corresponds to one
Target image block is provided with one first staging area and one second staging area.This first staging area has at least
One first random access point.This second staging area has at least one second random access point, independent of this at least one
First random access point.First, step S71 is that the multiple first kind encoded block in this target image block are applied
With image processing, to produce one group of first image processing result.Then, step S72 be to this group first image at
Reason result imposes frame buffering coding, to produce one group of first frame buffering coding result.Subsequently, step S73 is by this
Organize the first frame buffering coding result and be stored in this first staging area in this frame buffer.Step S74 is then by this mesh
At least one Equations of The Second Kind encoded block of mark image block is stored into this second staging area in this frame buffer.Must say
Bright, step S74 can exist independent of step S71～step S73, that is, its execution time can not be subject to step
The restriction of the execution time of S71～step S73.
Persond having ordinary knowledge in the technical field of the present invention it is understood that previously introduce image processing system 300,
Various may the change describing when 500 also can apply the image treatment method to Fig. 7, and its details repeats no more.
Idea of the invention can apply H.264, in the audio-visual compression coding technology such as high-effect image coding (HEVC), but
Not to be limited.
By the above detailed description of preferred embodiments it would be desirable to more clearly describe feature and the spirit of the present invention,
And not scope of the invention is any limitation as with above-mentioned disclosed preferred embodiment.On the contrary, its mesh
Be intended to cover various changes and tool equality the category being arranged in the scope of the claims that the present invention to be applied for
Interior.
Claims (20)
1. a kind of image processing system, in order to carry out image processing, the plurality of image area for multiple image block
Each image block in block comprises multiple first kind encoded block and at least one Equations of The Second Kind encoded block, at this image
Reason system comprises：
Multiple first kind encoded block in one target image block are imposed image processing by one image processing module,
To produce one group of first image processing result；
One frame buffering coding module, imposes frame buffering coding to this group the first image processing result, to produce one group the
One frame buffering coding result；
One frame buffer, corresponding to this target image block, is provided with one first staging area and one second working area
Domain, this first staging area has at least one first random access point, and this second staging area has at least one second
Random access point, independent of this at least one first random access point；And
One frame buffer controller, make this group first frame buffer coding result be stored into this frame buffer this first keep in
Region, and make this target image block at least one Equations of The Second Kind encoded block be stored into this frame buffer this second
Staging area.
2. image processing system as claimed in claim 1 is it is characterised in that this frame buffer controller makes this group
First frame buffering one first random access point from this at least one first random access point for the coding result is by continuous nothing
Compartment of terrain is stored in this first staging area.
3. image processing system as claimed in claim 1 is it is characterised in that this group first frame buffering coding is tied
Fruit comprises multiple frame buffering coding results, and this first staging area has multiple first random access points；This frame buffers
It is first temporary that controller makes the plurality of frame buffering coding result each be stored in this from the first different random access points
Deposit region.
4. image processing system as claimed in claim 1 is it is characterised in that this frame buffer controller makes this mesh
Mark image block this at least one Equations of The Second Kind encoded block from this at least one second random access point one second with
The continuous no compartment of terrain of machine access point is stored in this second staging area.
5. image processing system as claimed in claim 1, one, this target image block comprises many its feature
Individual Equations of The Second Kind encoded block, this second working area has multiple second random access points, and this frame buffer controller makes this
Multiple Equations of The Second Kind encoded block are each stored in this second staging area from the second different random access points.
6. image processing system as claimed in claim 1 is it is characterised in that correspond to adjacent to this target shadow
As a reference image block of block, this frame buffer is provided with one first with reference to staging area and one second with reference to temporarily
Deposit region, this first reference working area has at least one first random access point, this second reference staging area has
At least one second random access point independent of this at least one first random access point；This image processing module is to this mesh
When the plurality of first kind encoded block in mark image block imposes image processing, this frame buffer controller delays from this frame
This at least one second random access point rushing this second reference staging area of device reads in this reference image block
At least one Equations of The Second Kind encoded block, is supplied to this image processing module reference.
7. image processing system as claimed in claim 6 is it is characterised in that this image processing module is further
Image processing is carried out to this at least one Equations of The Second Kind encoded block of this reference image block being read, to produce one group
Second image processing result is so that this group the second image processing result is relevant to this group the first image processing result.
8. image processing system as claimed in claim 7 is it is characterised in that this frame buffering coding module is also right
This group the second image processing result imposes frame buffering coding, to produce one group of second frame buffering coding result；This frame delays
Rushing controller makes this group second frame buffer coding result being somebody's turn to do corresponding to this reference image block from this frame buffer
This at least one second random access point of second reference staging area starts to be subsequently saved back this frame buffer.
9. image processing system as claimed in claim 7 is it is characterised in that this image processing module root in the lump
According to the plurality of in this at least one Equations of The Second Kind encoded block of this reference image block and this target image block
One class encoded block, to produce this group the first image processing result and this group the second image processing result.
10. image processing system as claimed in claim 1 it is characterised in that this frame buffering coding module according to
One compression efficiency threshold value is compressed to this group the first image processing result to obtain this group first frame buffering coding
Result.
A kind of 11. image treatment methods, are carried out at an image in order to coordinate one frame buffer to be directed to multiple image block
Reason program, each image block in the plurality of image block comprises multiple first kind encoded block and at least one second
Class encoded block, this frame buffer is provided with one first staging area and one second temporary corresponding to a target image block
Deposit region, this first staging area has at least one first random access point, this second staging area has at least one
Second random access point, independent of this at least one first random access point, this image treatment method comprises：
A () imposes image processing to the multiple first kind encoded block in this target image block, to produce one group
One image processing result；
B () imposes frame buffering coding to this group the first image processing result, to produce one group of first frame buffering coding knot
Really；
C () makes this group first frame buffer this first staging area that coding result is stored into this frame buffer；And
(d) make at least one Equations of The Second Kind encoded block of this target image block be stored into this frame buffer this is second temporary
Deposit region.
12. image treatment methods as claimed in claim 11 are it is characterised in that step (c) comprises：Make this group
One frame buffering one first random access point from this at least one first random access point for the coding result is by continuously continuously
It is stored in this first staging area every ground.
13. image treatment methods as claimed in claim 11 are it is characterised in that this group first frame buffering coding is tied
Fruit comprises multiple frame buffering coding results, and this first staging area has multiple first random access points；Step (c)
Comprise：Make the plurality of frame buffering coding result each be stored in this from the first different random access points first to keep in
Region.
14. image treatment methods as claimed in claim 11 are it is characterised in that step (d) comprises：Make this target
This at least one Equations of The Second Kind encoded block of image block is one second random from this at least one second random access point
The continuous no compartment of terrain of access point is stored in this second staging area.
15. image treatment methods as described in claim 11 are it is characterised in that this target image block comprises
Multiple Equations of The Second Kind encoded block, this second working area has multiple second random access points, and step (d) comprises：Order
The plurality of Equations of The Second Kind encoded block is each stored in this second staging area from the second different random access points.
16. image treatment methods as claimed in claim 11 are it is characterised in that correspond to adjacent to this target shadow
As a reference image block of block, this frame buffer is provided with one first with reference to staging area and one second with reference to temporarily
Deposit region, this first reference working area has at least one first random access point, this second reference staging area has
At least one second random access point independent of this at least one first random access point；This image treatment method is in step
A taking a step forward of () comprises：
Read this reference from this at least one second random access point of this second reference staging area of this frame buffer
At least one Equations of The Second Kind encoded block in image block, is supplied to the image processing reference that step (a) is carried out.
17. image treatment methods as claimed in claim 16 are it is characterised in that comprise further：
Image processing is carried out to this at least one Equations of The Second Kind encoded block of this reference image block being read, to produce
One group of second image processing result is so that this group the second image processing result is relevant to this group the first image processing result.
18. image treatment methods as claimed in claim 17 are it is characterised in that comprise further：
This group the second image processing result is imposed with frame buffering coding, to produce one group of second frame buffering coding result；
And
Make this group second frame buffer coding result from this frame buffer correspond to this reference image block this second
This at least one second random access point with reference to staging area starts to be subsequently saved back this frame buffer.
19. image treatment methods as claimed in claim 17 are it is characterised in that step (a) comprises：Basis in the lump
In this at least one Equations of The Second Kind encoded block of this reference image block and this target image block the plurality of first
Class encoded block, to produce this group the first image processing result and this group the second image processing result.
20. image treatment methods as claimed in claim 11 are it is characterised in that step (b) comprises：According to a pressure
Contracting efficiency threshold value is compressed to this group the first image processing result to obtain this group first frame buffering coding result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510551785.6A CN106488240A (en) | 2015-09-01 | 2015-09-01 | Image processing system and image treatment method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510551785.6A CN106488240A (en) | 2015-09-01 | 2015-09-01 | Image processing system and image treatment method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106488240A true CN106488240A (en) | 2017-03-08 |
Family
ID=58237829
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510551785.6A Pending CN106488240A (en) | 2015-09-01 | 2015-09-01 | Image processing system and image treatment method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106488240A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101567849A (en) * | 2009-04-30 | 2009-10-28 | 炬才微电子（深圳）有限公司 | Data buffer caching method and device |
| CN101847394A (en) * | 2009-03-26 | 2010-09-29 | 珠海扬智电子有限公司 | Storage mapping method and device for encoding and displaying video files |
| CN101895775A (en) * | 2009-05-21 | 2010-11-24 | 华映视讯（吴江）有限公司 | Image processing method |
| US20120033738A1 (en) * | 2010-07-07 | 2012-02-09 | Steve Bakke | Virtual frame buffer system and method |
| CN103828370A (en) * | 2011-10-31 | 2014-05-28 | 联发科技股份有限公司 | Apparatus and method for buffering context arrays referenced for performing entropy decoding upon multi-tile encoded picture and related entropy decoder |
-
2015
- 2015-09-01 CN CN201510551785.6A patent/CN106488240A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101847394A (en) * | 2009-03-26 | 2010-09-29 | 珠海扬智电子有限公司 | Storage mapping method and device for encoding and displaying video files |
| CN101567849A (en) * | 2009-04-30 | 2009-10-28 | 炬才微电子（深圳）有限公司 | Data buffer caching method and device |
| CN101895775A (en) * | 2009-05-21 | 2010-11-24 | 华映视讯（吴江）有限公司 | Image processing method |
| US20120033738A1 (en) * | 2010-07-07 | 2012-02-09 | Steve Bakke | Virtual frame buffer system and method |
| CN103828370A (en) * | 2011-10-31 | 2014-05-28 | 联发科技股份有限公司 | Apparatus and method for buffering context arrays referenced for performing entropy decoding upon multi-tile encoded picture and related entropy decoder |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111434112B (en) | Method and device for acquiring global matching patch | |
| US9924199B2 (en) | Method and apparatus for compressing image data using compression profiles | |
| US5724451A (en) | Image compression method and apparatus using conditional quadtree split sysyem | |
| US5572691A (en) | Apparatus and method for providing multiple data streams from stored data using dual memory buffers | |
| CN1110965C (en) | Method and apparatus for efficient addressing of dram in a video decompression processor | |
| US20190082185A1 (en) | Efficient lossless compression of captured raw image information systems and methods | |
| US20140086309A1 (en) | Method and device for encoding and decoding an image | |
| CN101729840A (en) | Storage processing method utilizing video image characteristic sequence | |
| CN105472442A (en) | Out-chip buffer compression system for superhigh-definition frame rate up-conversion | |
| US8467624B2 (en) | Image processing apparatus, image processing circuit, and image processing method | |
| US11044466B2 (en) | Image processing device | |
| CN106488240A (en) | Image processing system and image treatment method | |
| US7386178B2 (en) | Method and apparatus for transforming the dimensions of an image | |
| WO2013011355A1 (en) | Method and apparatus for encoding an image | |
| JPH04326669A (en) | Picture encoding device | |
| TWI565303B (en) | Image processing system and image processing method | |
| CN107241601B (en) | Image data transmission method, device and terminal | |
| US6205251B1 (en) | Device and method for decompressing compressed video image | |
| CN112102144A (en) | Compressed data arrangement method and device and electronic equipment | |
| CN110876062A (en) | Electronic device for high-speed compression processing of feature map and control method thereof | |
| CN104581175A (en) | Image processing device and method | |
| US20100254618A1 (en) | Method for Accessing Image Data and Related Apparatus | |
| CN115278265A (en) | Parameter processing method, device, medium and electronic equipment | |
| CN100534189C (en) | Data preprocessing method and device | |
| CN112866723A (en) | Image data processing method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170308 |
|
| WD01 | Invention patent application deemed withdrawn after publication |