CN101472182B - Virtually lossless video data compression - Google Patents
Virtually lossless video data compression Download PDFInfo
- Publication number
- CN101472182B CN101472182B CN2008101861727A CN200810186172A CN101472182B CN 101472182 B CN101472182 B CN 101472182B CN 2008101861727 A CN2008101861727 A CN 2008101861727A CN 200810186172 A CN200810186172 A CN 200810186172A CN 101472182 B CN101472182 B CN 101472182B
- Authority
- CN
- China
- Prior art keywords
- compression
- video data
- video
- pixel
- compressed
- 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
Abstract
An apparatus and method of performing YUV (or YCrCb) video compression prior to storage within a memory and decompression upon retrieval of the blocks from memory. Compression is performed utilizing a quantizer to compress vide o data to a desired overall compression ratio R, even though the luma and chro ma contributions to compression can differ for each subblock, each preferably selected in response to texture estimation. Selections are made for each subblock to perform either linear or non-linear quantization during compression. Compression is performed without utilizing data from blocks outside of the block being compressed, wherein video blocks can be retrieved and decompressed in any desired order. In one implementation, an encoder non-sequentially selects blocks from memory which are then decompressed and encoded. The compression may be beneficially utilized in a number of different video transmission and storage applications without departing from the invention.
Description
Technical field
Relate generally to Video processing of the present invention, more specifically, the present invention relates to encode or export before compression and the decompress(ion) of video in video memory.
Background technology
Implement typical video framework, so that Video Data Storage was being carried out preliminary treatment to it before the video frame memory that usually visits by external bus.Video data is taken out from video memory and is exported during coding is processed subsequently, perhaps takes out from video memory so that output.The quantity that the data of being undertaken by bus between encoder and video memory are transmitted may be very large, needing to cause the multitude of video memory, requires huge memory bus bandwidth, and result causes very high power consumption.
Fig. 1 illustrates the framework of traditional video camera (video camera) or static images camera (still-picture camera).So the input video that obtains by the entering apparatus such as CCD or cmos sensor is stored in external memory storage after preliminary treatment.Usually, preliminary treatment comprises noise reduction and the video data conversion of (brightness (luminance) Y, two colourities (chrominance) U+V) from RGB (RGB) to YUV.Other treatment facilities such as video encoder or display device are reading video data from video memory subsequently.
Therefore because the video data volume is generally larger, carrying out video data by external bus, to transmit required bus bandwidth quite high.Particularly for HD (high definition) Video Applications, required bus bandwidth is very huge, and wherein bandwidth cost and power consumption become very high, makes to realize that cheaply the image/video system is very difficult.
Another technological challenge in video camera is relevant with required memory size.Most of video parts are to realize with the form of Soc (SOC (system on a chip)).Usually, the cost of external memory storage (as SDRAM) is usually above other devices.Therefore, required memory size reduce can make this reduction of system synthesis.
Therefore, need a kind of reducing for the necessary memory bandwidth of video data stream and the method for memory, have this needs when particularly carrying out the method before Video coding is processed.The present invention has realized that this needs and other need and overcome the restriction of previous solution.
Summary of the invention
The apparatus and method that are used for the virtually lossless video data compression that adopt yuv video data compression/decompression method have been described.For simplicity, adopt the term yuv video here, should be appreciated that simultaneously, instruction and claim that YCrCb video or other similar codings are equally applicable to record and narrate here.Come the compressed video input by each frame being divided into the one group of compressed piece (compressedblock) that has a given number pixel (for example, being N pixel for Y, is M pixel for U, V) for each component.Therefore compression blocks (compression block) is not overlapping with his compression blocks and can be independently of one another processed and need not reference information from other compression blocks, wherein said video data can be accessed randomly, and this point is particularly suited for being used by video encoder.As a result, in view of the compression blocks size is less, in memory, the storage of compression blocks needs less memory stores and less bus bandwidth.Afterwards, compressed video data is removed from memory by the order (for example, sequentially non-) of any expectation and for example by the encoder decompress(ion), and/or be output to display (this is not too preferred).Decompress(ion) returns to data its unprocessed form, wherein utilizes the equipment of the data (decompressed data) through decompress(ion) or application to need not to know once compressed mistake of these data.
Preferably carrying out complexity before compression estimates in order to can determine optimal compression rank and pattern.
Between compression period, predict to predict current pixel value based on first preceding pixel, then determine the sub-block size and cut apart sub-block.Then, make decision to quantizing type (for example, linear or non-linear), this decision is for example preferably carried out for each sub-block.For nonlinear quantization, one aspect of the present invention provides the estimation of quantization parameter (QP), is not wherein that all possible QP value all needs searched.Quantize based on above-mentioned definite execution subsequently, the sub-block of compressed data is packed in piece the most at last.Between compression period, at first one aspect of the present invention carries out the colourity compression, wherein can obtained and employing during luminance compression from the information of colourity compression.In response to complexity and the available bits number of video data, aspects more of the present invention provide the selection to different compression levels.Compressed brightness and colourity are incorporated in final compressed piece.
Data complexity determines because the type that quantizes and rank are in response to, and compressed video data after decompress(ion) (although Incomplete matching original video data) similarly is still virtually lossless in human viewer.Therefore, apparatus and method of the present invention generate output desired qualities, that be equivalent to the virtually lossless quality simultaneously in the situation that memory requirement reduces and bandwidth reduces to work.
The target that reduces bus bandwidth and/or memory size provides in the above as example.The application of compression method is not limited to this concrete application advantage, but can be used to provide advantage in any other application facet that needs video data compression, for example pass through perhaps before decompress(ion), compressed Video Data Storage to be arrived aspect any other medium apparatus such as hard disk or memory aspect the compressed video data of Internet Transmission before decompress(ion).
The present invention can be modified to realize in several ways, include but not limited to following description.
one embodiment of the present invention are the video memory device that are configured to carry out with one or more video modules the transmission of video pixel data, it comprises: the video compressing module that (a) is coupled to described video memory by signal bus, be configured to (a) (i) in the situation that be not used to from the pixel data of each piece with exterior pixel, use quantizer brightness and colourity YUV inputting video data to be compressed into the compressed video data of the every pixel bit number with minimizing, (a) (ii) with compressed Video Data Storage in video memory, and (b) video decompression module, be configured to take out compressed block of video data and the compressed video data solution of storing in video memory is pressed into by any order have same format and approximately received and the video data through decompress(ion) of the original video data compressed by video compressing module, wherein said video decompression module is configured to export the video data through decompress(ion).
In video-unit, to compress and preferably carry out according to total draught R, the degree of total draught R controll block compression also can represent by the mode (ratiometrically) of ratio metering or according to the bit number that comprises in resulting.In at least one preferred embodiment, be that brightness and colourity are selected identical or different compression level when keeping total draught R.In at least a optimal way, unless colourity has high levels of complexity, otherwise the degree of compression is by this Array selection, to reach the MIN bit that brightness is used of optimizing simultaneously at the bit that colourity is used.
In at least a execution mode of this video-unit, estimate texture (texture) complexity for compression blocks before compression.The processing that the interior compression stroke to brightness and colourity of each piece distributes (being for example the bit number that brightness and colourity are distributed in compressed piece) to be in response to the estimation Texture complication is determined.
In at least a execution mode, brightness and chrominance information are incorporated in compressed block of video data, and compressed block of video data is preferably filled to keep the fixed size of compressed piece with filling bit.
In at least a configuration of this device, the luminance compression of given is adopted the information of determining during the colourity compression of this piece is processed.In at least a mode of the present invention, video compression is for example to utilize the nonlinear quantization in the combination of linear and nonlinear quantization to carry out.In a preferred embodiment, when carrying out nonlinear quantization, adopt the quantization step of different accuracy.
In at least a mode of the present invention, carry out pixel prediction between compression period.Pixel prediction starts from having defined in the central initial reference pixel of selecting of piece the right prediction direction and the left prediction direction that begin from this pixel, and their can be processed on demand side by side.Should be appreciated that, by selecting center pixel (or adjacent central) to begin prediction, so reference pixel both keeps identical to right prediction direction and left prediction direction.The prediction of pixel value is by carrying out based on previous pixel value prediction current pixel value, and the prediction of both direction therebetween is independently, thus therefore allow right and left to prediction processing carry out side by side reducing the required processing time.
In at least a execution mode, sub-block is cut apart and is to carry out according to the configuration of expectation.In order to determine that sub-block cuts apart, at first a plurality of at least a portion that may the sub-blocks configuration are calculated the sub-block value at cost, abandon afterwards the sub-block configuration that cost surpasses available bit number in given cost threshold value and/or compressed piece.
In at least one embodiment of the present invention, the inputting video data for device is to receive from the imageing sensor that for example may be integrated in video camera or still camera or other imaging devices.The form of inputting video data is configured to not only comprise monochrome information but also comprise chrominance information.
According at least one target of the present invention, compression and decompress(ion) are to carry out towards the direction that reduces bus bandwidth and/or video memory demand, for example as the pretreated a kind of form of yuv video data of carrying out before the coding of following, described be distinctive with encoder be that the mode of the non-order order of piece number (for example, not according to) is taken out from memory usually.As a supplement or substitute, can implement according to compression of the present invention and decompress(ion) when sending compressed video data by network before decompress(ion), perhaps implement according to compression of the present invention and decompress(ion) in case before decompress(ion) with Video Data Storage in medium apparatus.
One embodiment of the present of invention are that this device comprises for the video encoder apparatus of coding yuv video data: (a) video memory, and this video memory is configured to carry out with one or more video modules the transmission of video pixel data; (b) be coupled to the video compressing module of video memory by signal bus, be configured to (b) (i) in the situation that need not reference data from other pieces, use quantizer brightness and colourity YUV inputting video data to be compressed into the compressed video data of the every pixel bit number with minimizing; (b) (ii) with compressed Video Data Storage in video memory; (c) video decompression module is configured to take out piece and will compressed video data decompress(ion) be have same format and approximately received and the video data through decompress(ion) of the original video data compressed by video compressing module from video memory by the order of any expectation; And (d) coding module, non-ly sequentially select block of video data, receive through the video data of decompress(ion) and it is encoded from memory.
One embodiment of the present of invention are methods of compression and decompress(ion) yuv video data, comprise that (a) uses quantizer by compression ratio R, each piece that inputting video data is compressed into for video data to be had the brightness of minimizing and/or the compressed block of video data of chrominance bit number; (b) wherein the compression of inputting video data is in the situation that be not used to carry out from the data of the pixel of compressed outside; (c) with compressed Video Data Storage in video memory; And (d) in order or any block of video data of selecting of the order of non-order, with compressed video data decompress(ion) to generate the video data output through decompress(ion).In one embodiment, compression or decompress(ion) are in conjunction with the compression of video data is processed to carry out, wherein selected from the non-order piece of the video data of memory, be extracted according to the present invention and be received for coding.
One embodiment of the present of invention are methods of compression and decompress(ion) yuv video data, comprise that (a) uses quantizer by compression ratio R, each piece that inputting video data is compressed into for video data to be had the brightness of minimizing and/or the compressed block of video data of chrominance bit number; (b) in the situation that be not used to the compression each video block carried out from the data of compressed outside; (c) each sub-block that will be performed in given of compression is selected linearity or nonlinear quantization; (d) with compressed Video Data Storage in video memory; And (e) for any block of video data from the video memory taking-up of selecting by any order, compressed video data decompress(ion) is exported to generate through the video data of decompress(ion).
One embodiment of the present of invention are methods of compression and decompress(ion) yuv video data, comprise that (a) uses quantizer by compression ratio R, each piece that inputting video data is compressed into for video data to be had the brightness of minimizing and/or the compressed block of video data of chrominance bit number; (b) compression is in the situation that be not used to from the data of the piece of compressed outside, block of video data to be carried out; (c) estimated brightness Texture complication and chrominance texture complexity; (d) select identical or different compression level when keeping total draught R in brightness and colourity; (e) in response to the characteristic that detects, each sub-block that be performed in given of compression is selected linearity or nonlinear quantization in piece; (f) wherein between compression period, the luminance compression processing and utilizing is from the information that the colourity compression of same block is processed; (g) with compressed Video Data Storage in video memory; (h) any moment after the compressed video data of storage, take out block of video data by the order of any expectation from video memory; (i) for the piece that takes out, compressed video data decompress(ion) is exported to generate through the video data of decompress(ion).
The invention provides a plurality of favourable aspects, they both can separately realize also can realizing by the combination of any expectation, only otherwise breaking away from this instruction gets final product.
One aspect of the present invention comprises for the compression of yuv video data block and the apparatus and method of decompress(ion).
Another aspect of the present invention is that each compressed piece other compressed pieces of getting along well are overlapping, and each compression blocks is in the situation that with reference to being compressed independently from the information of other pieces.
Another aspect of the present invention is, if the target bits that after compression, the bit number of compressed information matches less than the compression ratio (R) with expectation extends to the number of expectation so that this number keeps fixing with filling bit with always exporting bit.Because the size of the number of the bit of each compression blocks that generates is fixed, the compressed piece that therefore is in any position can accessed and decompress(ion), and need not the information with reference to other pieces.So, use the method propose, to can easily carry out from the access of the video data of random site in video and the decompress(ion) of video block, for example the arbitrary region in frame needs accessed in order to carry out the situation of Video processing or coding.
Another aspect of the present invention is that brightness data is by using the information of determining as the result of colourity compression to compress.In the method according to this invention, compressed chromaticity at first.After the colourity compression, it is that luminance compression is prepared that at least a portion of this information is passed to brightness module.These information bits can be retained time period or the bit number of any expectation after compression.Luminance compression is utilized this information when compression.Finally, compressed brightness and chroma data are packaged advances in a compression blocks.Size of data is guaranteed within the scope of target compression appointment.
Another aspect of the present invention is the Texture complication of estimated brightness and colourity compression blocks at first before compression.Each compression module (brightness and colourity) can automatic lifting for different compression levels (degree) (for example, a little less than, in, by force).The degree of compression is based on Texture complication to be determined, is wherein brightness and colourity field allocation bit number in compressed piece subsequently.
Another aspect of the present invention is to the utilization by the given total draught of R, and R controls the compressed total degree of video block.For given piece, the value of R is illustrated as ratio (input bit number/output bit number).Ratio R is two (2) to show that compressed piece contains half of bit number that initial data adopts.Yet, should be appreciated that, compression can represent by the form of any expectation, for example any type of ratio is specified, perhaps with select given every video IOB bit number relevant.
Another aspect of the present invention allows to select Y, U and the compressed degree of V when adhering to total draught R.Should be noted that this does not mean that for each compression blocks, between Y, U and V, compression ratio is fixed, because can come allocation bit based on the different compression ratios between colourity and brightness.
Another aspect of the present invention is automatically to determine the processing of the optimal compression value between brightness and colourity when keeping total draught R.Substantially, human eye is more responsive to the signal noise of the signal noise ratio colourity of brightness.Therefore, the bit that preferred Bit Allocation in Discrete strategy uses colourity reaches bottom line, and the bit that brightness is used reaches maximum.Yet when finding that colourity has a certain levels of complexity, particularly when finding that colourity is very complicated, at least one embodiment of the present invention allows colourity to prevent the appearance of visual artifact with enough bit numbers.This maximization that bit is used and minimizing is in response to that the information about brightness and chrominance texture complexity that obtains determines.
Another aspect of the present invention is selected different compression levels.For example, can define low, high-quality compact model neutralizes.The low quality pattern is based on adopting high compression rate, and high quality mode adopts little compressible.According to an aspect of the present invention, final optimal compression pattern is determined.Should be noted that and can both define compression level to luminance compression and colourity compression, wherein the part of the piece part of brightness use and colourity use is determined.
Another aspect of the present invention is to determine that the decision of optimal compression pattern is that the information about Texture complication and available bits of passing through to obtain is made.
Another aspect of the present invention is from intermediate pixel rather than the prediction of first pixel starting pixel of a string pixel, and wherein other pixels are predicted relatively with this initial pixel.Two prediction direction are calculated, and wherein the required processing time can reduce.
Another aspect of the present invention is in response to and quantizes and the sub-block of sub-block configuration decision is cut apart.Total draught R is in response to for the pixel count N of brightness and is used for that the pixel count M of colourity compression (U and V, or Cr and Cb) determines.For given brightness and colourity compressed value (N, M), there are some different sub-blocks configurations.Sub-block size decision module or similar module determine that according to given one group of input and condition optimum sub-block configures.
Another aspect of the present invention be based on given information come at least a portion of different possible sub-block configurations (for example, perhaps all different sub-blocks combination) assess the cost, described information is such as being the information that receives from from the information (Information_from_chroma) of colourity, R, QP precision etc.For instance, cost can be estimated according to the number of the output bit that generates.If the output bit number of given sub-block configuration surpasses the threshold value value at cost that preferably provides by available bit number, abandon this sub-block configuration and check next possible sub-block configuration.In at least one execution mode, if do not have possible configuration available, the original compression piece of N pixel is used as sub-block.
Another aspect of the present invention is to select linearity or nonlinear quantization between the piece compression period.
Another aspect of the present invention is for the estimation for the quantization parameter (QP) of nonlinear quantization described herein.
Other aspects of the present invention will occur in the following part of specification, and wherein describing in detail is for comprehensive openly the preferred embodiments of the present invention, rather than it is applied restriction.
Description of drawings
By with reference to the following accompanying drawing that only is used for the illustrative purpose, can comprehend the present invention:
Fig. 1 is the block diagram of conventional video camera or still camera framework, and the storage of pretreated view data in video memory is shown;
Fig. 2 be according to an aspect of the present invention video camera or the block diagram of still camera framework, be illustrated in video memory the compression of view data before storage;
Fig. 3 is the block diagram of compressed according to an aspect of the present invention;
Fig. 4 is the block diagram of piece decompress(ion) according to an aspect of the present invention;
Fig. 5 be according to an aspect of the present invention colourity and the flow chart of luminance compression, be illustrated in the information from the colourity compression that luminance compression adopts in processing;
Fig. 6 is total compression flow chart according to an aspect of the present invention;
Fig. 7 is the flow chart of supporting according to an aspect of the present invention different compact models;
Fig. 8 makes the flow chart that compression level determines according to an aspect of the present invention, is illustrated in when making a decision compression level and carrying out the preparation of compression method estimation and analysis to Texture complication;
Fig. 9 is the flow chart of compression method according to an aspect of the present invention, and the linear or non-linear Q that piece is carried out in response to quantizing decision is shown;
Figure 10 is the pixel map of pixel prediction according to an aspect of the present invention, and the selection to the reference pixel of a plurality of pixel central authorities is shown;
Figure 11 is in response to the selection of as shown in figure 10 reference pixel, pixel map on right prediction direction and left prediction direction;
Figure 12 is the block diagram of chooser block size according to an aspect of the present invention;
Figure 13 is the flow chart of determining according to an aspect of the present invention the sub-block size;
Figure 14 is according to an aspect of the present invention in response to the flow chart of the nonlinear quantization of the QP value of estimating and residual error (residue) data;
Figure 15 is the flow chart that QP according to an aspect of the present invention estimates, is illustrated as carrying out in response to definite residual error data.
Figure 16 is the block diagram of selecting according to an aspect of the present invention linearity or nonlinear quantization;
Figure 17 is the flow chart that determines according to an aspect of the present invention linearity or nonlinear quantization.
Embodiment
More specifically with reference to accompanying drawing, the present invention is embodied as generally by Fig. 2 for illustration purposes to device shown in Figure 17.Should be appreciated that, under the prerequisite that does not break away from basic conception disclosed herein, this device just can change with regard to configuration and with regard to the details of some part, and this method can change with regard to concrete steps and order.
[yuv video data compression/decompression method]
Fig. 2 illustration be used for to carry out the compression of YUV (or YCrCb) video data and the embodiments of the invention 10 of decompress(ion).Can utilize compression to present the advantage of any number relevant to given application, in order to for example reduce bus bandwidth and video memory demand.
Apparatus and method of the present invention receive input (12) from video equipment, deal with data (14), and then compressed video data (16), then for example be stored in external video memory 20 by bus 18.Video encoder 24 or display device 28 take out compressed video and with its decompress(ion) (22,26), utilize afterwards this video data from video memory.Because video data compressed before being stored in video memory, therefore required bus bandwidth is more much lower than originally, and the necessary size of video memory (for example SDRAM) also reduces similarly simultaneously.
Fig. 3 illustration compressed YUV or similar according to the present invention example embodiment 30.The inputting video data form of the compression blocks 32 of compression method of the present invention is YUV (or YCrCb).During compression was processed, frame (video data by Y, Cr and Cb consists of in this example) was divided into one group of compression blocks.Compression blocks is defined as for N pixel of brightness (Y) 34 and is used for each M pixel of colourity (Cr, Cb) 26,38, as shown in the figure.Therefore, the total pixel number in each compression blocks is (N+2M).If adopt B bit to represent a pixel, total bit number is (N+2M) * B bit.
Compression blocks other compression blocks of getting along well are overlapping, and each compression blocks is compressed and independently not with reference to the information in other pieces.
In case compression ratio R is received for input, an execution mode of compression method 40 just generates the compressed piece 42 of fixed size.The number that generates bit is calculated by following formula:
Total output bit=(N+2M) * B/R
For example, if R is 2, half that the total output bit that generates is original YUV bit number.
If after compression, the bit number of compressed information is less than the target bits by the R appointment, preferably bit interleaving is kept fixing total output bit number all the time.Because the size of the bit of each compression blocks that generates is fixed, the compressed piece that therefore is in any position all can be in the situation that be acquired and decompress(ion) with reference to information in other pieces.Therefore, using this inventive method, is possible to the access of the video data of the random site of (as the compression blocks unit).This needs accessed necessary in order to carry out the application of other Video processing for arbitrary region in frame.
Fig. 4 has described the embodiment 50 of decompression processing.Decompressing method is accepted compression blocks 52, and compression blocks 52 is extracted (54) to returning to the original YUV resolution compression blocks 56 of (for example, N pixel is used for Y 58, and M pixel is used for U 60, and M pixel is used for V).Yuv data through decompress(ion) also can be utilized in other equipment in order to carry out Video processing.Because the dimension of decompress(ion) rear video data is identical with initial data, so other treatment facilities can't be recognized any difference of being introduced by compression algorithm based on video format.
The embodiment 70 of Fig. 5 illustration piece compression and the relation between brightness and colourity compression module.In at least a execution mode, luminance compression is in response to that the information received carries out between the colourity compression period to same block.Can find out, at first receive colourity (UV) data 72 for colourity compression 74, wherein the information information_from_chroma from the colourity compression is passed and is used for carrying out luminance compression 78 in conjunction with brightness (Y) data 76.It is therefore preferable that according to the inventive method and at first carry out the colourity compression.Should be appreciated that, can keep time or the bit number of any expectation from the information of colourity after compression, for example in luminance compression.Finally, as shown in piece 80, compressed brightness and chroma data are packaged to be advanced in a compressed piece 82.According at least a preferred implementation, size of data is guaranteed in the scope by the target compression appointment.
The embodiment 90 of the invention compression method that Fig. 6 illustration is total.The reception of colourity compression blocks 92 and luminance compression piece 94 is illustrated.Before compression, the Texture complication of colourity is estimated that the Texture complication of (96) and brightness estimated (100) (not hinting concrete order).Colourity compression 98 and luminance compression 102 are to utilize from brightness and colourity complexity to estimate that both the data of reception are carried out, and luminance compression is also from colourity compression reception information simultaneously.Compressed brightness and chroma data be received and packing (104) subsequently, and compressed piece 106 generates.
Considering compression is processed back, should be appreciated that, each compression module of the present invention (brightness and colourity) can provide different compression level (degree) (for example, a little less than, in, by force).According at least a execution mode of the present invention, the degree of compression is preferably selected in response to the rank of Texture complication.
By the rate control Y that variable R represents, the compression ratio of U and V compression blocks.Should be noted that from this point and can not infer Y, the fixedly compression of U and V experience R; Because Y, U and V (or Y, Cr and Cb) can be compressed to respectively separately the ratio of any expectation, as long as total draught R is followed in resulting compression.Therefore, can come allocation bit based on the different compression ratios between colourity and brightness.
In one embodiment of the invention, the method that realizes attempts optimizing the compression ratio (that is, seeking optimal compression value in the restriction of technology and available information) between brightness and colourity when total draught is remained R.This aspect of the present invention with human eye to more responsive the taking into account of the general chrominance noise of brightness noise.Therefore, in preferred Bit Allocation in Discrete strategy, colourity adopts MIN bit number so that the bit that brightness is used can maximize.Yet, in some cases, for example find very complicated situation of colourity, pattern of the present invention allows colourity to adopt enough bit numbers to prevent visual artifact (visual artifact).
In at least a execution mode, the Bit Allocation in Discrete ratio is to select on the basis of determining brightness and chrominance texture complexity.For example calculate compression blocks residual error on average can obtain Texture complication in Fig. 6.
The embodiment 110 of inventive method of the different compression quality patterns of compression level is depended in the support of Fig. 7 illustration.For example, according to the present invention, compression blocks 112 can experience low (114), in (116) and high-quality (118) compact model.The low quality mode producing is in the use of high compression rate, and the high-quality video pattern comes from the use of little compressible.In a kind of preferred implementation, operation processes 120 selecting final optimal compression pattern, thereby produces final compressed piece 122.Should be realized that, compression level can define in both at brightness and colourity compact model.
An example implementing the compression of different stage (pattern) is by adopting the bit accuracy of output quantization.For low compact model, can adopt higher precision, and can adopt lower precision to high compression mode.Compression ratio according to expectation can pre-determine quantified precision on different stage.As shown in Figure 7, the optimal compression pattern determines to be generated in the information of relevant Texture complication and available bits by system responses.
Fig. 8 illustration is made the example embodiment 130 that compression level determines according to the present invention.Brightness (Y) data 132 and colourity (UV) data 134 estimate that for the estimation 136 of luminance texture complexity and chrominance texture complexity 138 is received respectively.These complexities are estimated to describe in following paragraph.Brightness and chrominance texture are estimated to be utilized by texture analyzer 140 subsequently, the output of texture analyzer 140 is used to carry out compression level in piece and determines 142 and carry out subsequently compression method 144 by piece, to export compressed chroma data 146 and from the information 148 of colourity.Compression method is described in Fig. 5, and texture analysis and compression level determine to be described below.
[estimation of luminance texture complexity]
Following false code by example and non-limited way illustration the estimation of luminance texture complexity.For each sub-block, analyze the luminance texture complexity and estimate that complexity is as follows.
Index x: location of pixels index
Index bp: sub-block location index
Luma_Texture_Complexity(LTC)
The complexity of the current luminance compression piece of this value indication.Can define three different levels of complexity: a little less than, in, strong.
Detection method:
For each sub-block, calculate Residual[x]=Current pixel[x]-predicted
pixel[x]；
For each sub-block, seek Residual[x in sub-block] maximum, for example
max_Residual[bp]；
Calculate max_Residual[bp on whole compression blocks] average_residue (average residual
Poor);
If average_residue＜Threshold_1:
Whole luminance block is regarded as " weak complexity ".
If average_residue 〉=Threshold_2:
Calculate max_Residue[bp]＞number of the sub-block of Threshold_3;
If the number＜Threshold_4 that calculates:
Whole luminance block is regarded as " middle complex block "
Otherwise whole luminance block is regarded as " strong complex block ".
[estimation of chrominance texture complexity]
Following false code by example and non-limited way illustration the estimation of chrominance texture complexity.For each sub-block, analyze the chrominance texture complexity and estimate that complexity is as follows
Chroma Texture Complexity(CTC)
Detection method:
For each sub-block, calculate Residual[x]=Current pixel[x]-predicted
pixel[x]；
If any Residual[x]＜Threshold_1:
If the residual values of sub-block boundary is greater than Threshold_2:
Sub-block is regarded as " low complicated ".
Otherwise 4 * 1 colourity sub-blocks are regarded as " high complicated ".
[compression level decision]
The non-limited way by example can be implemented compression level according to following false code and determine.
If(CTC＝＝high)
If(LTC＝＝low)
Chroma_Compression_Level＝LOW_COMPRES_SION_MODE
Luma_Compression_Level＝HIGH_COMPRESSION_MODE
Else_if(LTC＝＝medium)
Chroma_Compression_Level＝HIGH_COMPRESSION_MODE
Luma_Compression_Level＝MIDDLE_COMPRES_SION_MODE
Else(LTC＝＝high)
Chroma_Compression_Level＝LOW_COMPRESSION_MODE
Luma_Compression_Level＝HIGH_COMPRES_SION_MODE
Else if(CTC＝＝low)
If(LTC＝＝low)
Chroma_Compression_Level＝LOW_COMPRESSION_MODE
Luma_Compression_Level＝HIGH_COMPRESSION_MODE
Else if(LTC＝＝medium)
Chroma_Compression_Level＝MIDDLE_COMPRESSION_MODE
Luma_Compression_Level＝MIDDLE_COMPRESSION_MODE
Else(LTC＝＝high)
Chroma_Compression_Level＝HIGH_COMPRESSION_MODE
Luma_Compression_Level＝MIDDLE_COMPRESSION_MODE
[each other compression method of level]
The example embodiment 150 of Fig. 9 illustration invention compression method.Compression blocks 152 is received by the prediction processing 154 of predicting current pixel value based on first preceding pixel.In case obtain predicted value, just calculate residual values (current pixel and predicted pixel poor) in prediction processing 154.The sub-block size determines to process 156 and determines optimum sub-block size based on Texture complication and available bits budget.In case determined the sub-block size in 156, the residual error data collection just is divided into the array of the less size that is called as sub-block in piece 158.Quantification treatment 160 is applied to each sub-block of residual error data to reduce the output bit.Therefore, each sub-block can use unique QP (quantization parameter) value to quantize.Accompanying drawing illustrates carries out equal interval quantizing 164, or estimates that the decision that then quantization parameter (QP) 162 carries out nonlinear quantization 166 is made.In arbitrary situation, the sub-block result all packaged (168) to produce final compressed piece 170.
Prediction means that current pixel is by previous pixel prediction.Predicted pixel is used for calculating " residual error " between actual current pixel value and predicted value.In the methods of the invention, residual error is quantized and encodes in follow-up phase.
X[n] prediction=F (X[n-1], X[n-2] ...)
X[n] residual error=abs (X[n]-X[n] prediction)
Figure 10 and Figure 11 difference illustration are used for initial pixel prediction 190 and are used for supporting the mechanism of two prediction direction 210.Should be appreciated that, pixel data set can be any size, and non-limited way illustrates the sample data collection by example for simplicity.In order to illustrate for simplicity, the describing one dimension data set is although these mechanism also can be applicable to 2-D data.In these figure, the position of pixel in the numeral indication compression blocks in square.
Should be noted that in Figure 10, prediction 190 is that center pixel 194 rather than first pixel from data 192 begins.The pixel 194 that is positioned at approximate center position (for example, position 16) is set as the initial reference pixel, and every other pixel all begins predicted from this pixel.As shown in figure 11, mechanism 210 is supported as right 212 and left to 214 two prediction direction that provide.Because reference pixel (for example, the pixel at 16 places, position) does not change, so in this both direction, the prediction of each is independently.Therefore, can process side by side right and left to prediction, the required processing time can greatly reduce thus.
Compression blocks is divided into sub-block.The purpose that sub-block is cut apart is to distribute independently QP for each sub-block.Because lower QP can be used to minimize the noise that causes by quantizing, if therefore be divided into some less sub-blocks, gross mass can be improved.Yet if the sub-block number is large, the expense of QP value increases.
[sub-block that is used for quantizing is cut apart and sub-block configuration decision]
There are a lot of utilizable may sub-block configurations in given (N, M) individual pixel." the sub-block size determines " module responds is determined best sub-block configuration in the parameter of receiving.
The example that sub-block is cut apart is based on the information that obtains after colourity compression.According to an aspect of the present invention, at first carry out the colourity compression, wherein the information from this processing can be provided to for during luminance compression.For example, total bit number and the chrominance texture complexity used in the colourity compression can be regarded as information_from_chroma.
Figure 12 illustration is made the embodiment 230 that the sub-block size determines in response to the reception of information_from_chroma, R and QP_precision in sub-block size decision module 232.The output of module 232 comprises best sub-block size, as 2 * 2, and 4 * 4,8 * 8 etc.
A possibility execution mode 250 as 13 illustrations " the sub-block size determines " module.This decision module receives information_from_chroma, R and the QP_precision that (252) are used for each possibility sub-block configuration.Be that each configuration (254,260,266) assesses the cost based on given information.Value at cost can be expressed as the estimated number of the output bit of generation.Determine that layer (ladder) (256,262) are illustrated, if wherein the output bit number is greater than available output bit number, abandons this customized configuration and next possible configuration is carried out inspection.Optional output is illustrated as 2 * 2 sub-block 258,4 * 4 sub-block 264 grades here until do not cut apart (N sub-block) output 266.Can find out, when not having possible configuration to accept, the original compression piece of N pixel is used as 266 and does not cut apart sub-block 268.
Should be appreciated that, inventive method is configured to support two kinds of quantification treatment, linearity and nonlinear quantizations.
[1. the quantization parameter in nonlinear quantization is estimated]
Given may QP the set of value, system can check that all QP values seek suitable selection.Yet, to the computation complexity inefficiency of the exhaustive search of QP and realize that cost is high.
Figure 14 be illustrated in do not search for might QP value prerequisite under estimation QP value reduction the embodiment 270 of method of complexity.Residual error data 272 is illustrated as being received by QP estimator 274.Residual error data and the QP value of determining by estimation are all processed 276 by non-linear Q and are received.
The embodiment 290 that the QP that Figure 15 illustration is shown specifically estimates.The sub-block residual error data 292 that adopts during for the maximum residul difference value (294) in seeking sub-block, QP estimator of the present invention uses original pixel value.Maximum is used when determining to quantize to determine rank (296) subsequently.In other words, can find the maximum rank (maximum decision level) that determines by the maximum residul difference value, determine that by maximum rank can determine (298) best Q P value.Table 1 illustrates the maximum example that determines rank and QP value of mapping.
Table 1
The maximum example mappings that determines rank and QP value
| DL | D0 | D1 | D2 | D3 | D4 | D5 | D6 | D7 |
| MRR | r＜4 | 4≤r＜8 | 8≤r＜16 | 16≤r＜32 | 32≤r＜48 | 48≤r＜64 | 64≤r＜128 | 128≤r＜256 |
| |
0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
DL-determines rank; MRR--maximum residul difference scope; The QP that E.QP-estimates
In order to determine the QP value of dibit Q nonlinear quantization, use the maximum of the residual error data of initial data to be used (QP estimation) in 4 * 1.In table 1, can find out the mapping between the QP of maximum residul difference (max_residual) data and estimation in sub-block.
[2. linear/non-linear quantizes to determine]
According to this aspect of the present invention, each sub-block can be according to the inventive method, be quantized in response to the selection of linearity or nonlinear quantization.For example, in response to the large residual values of sub-block, the QP value is also often larger, and wherein serious quantizing noise is introduced into, and causes producing unwanted visual artifact.
Figure 16 illustration for fear of the visual artifact such as noise produce, the embodiment 310 of or equal interval quantizing non-linear in response to the content choice of video data.In figure, be shown in the example parameter that adopts when selecting and be depicted as sub-block residual error 312 and raw pixel data 314.Yet, this stage can with any other parameter can use in quantizing to determine 316 equally, determine 318 to arrive final quantization method.
The embodiment 330 of Figure 17 illustration " non-linear/equal interval quantizing determines " module.Data such as sub-block residual error 332 and original sub-block pixel data value 334 are interior received in rim detection module 336.Should be appreciated that, rim detection module 336 can be implemented by the alternative with any number of original pixels or residual error or similar value.For example, if the maximum residul difference in sub-block greater than threshold value, can be thought strong marginal existence.If strong edge do not detected according to step 338, make the decision of using nonlinear quantization 346.Otherwise estimate (340) cost to use equal interval quantizing, cost compares to draw with available bit number the decision of using equal interval quantizing 344 subsequently in 342.If cost greater than available bits, is selected nonlinear quantization 346.
Although above description comprises a lot of details, it should not be construed as limiting the scope of the invention, and is interpreted as only providing to some explanation in the present preferred embodiment of the present invention.Therefore, should be appreciated that, scope of the present invention comprises other embodiment that may it will be apparent to those skilled in the art fully, and therefore scope of the present invention is only limited by claims, unless wherein done clearly statement, otherwise mentioned that with singulative certain key element do not want to refer to " one and only one " and refer to " one or more ".On all structures of the key element of above preferred embodiment known to persons of ordinary skill in the art and the equivalent on function clearly be incorporated into by reference this, and wish to be contained by current claim.In addition, a kind of equipment or method needn't solve each problem of attempting to be solved by the present invention in order to be included in current claim.In addition, any key element, assembly or the method step in present disclosure do not plan to dedicate to the public, and no matter whether these key elements, assembly or method step are are clearly recorded and narrated in the claims.Any claim here should not understood according to the regulation of the 6th section of 35U.S.C.112, unless use phrase " be used for ... device " recorded and narrated clearly this element.
The application requires the priority of the U.S. Provisional Application sequence number 61/015,174 submitted on December 19th, 2007, and the full content of this application is incorporated into this by reference.
Claims (25)
1. device that is used for compression and decompress(ion) yuv video data comprises:
Video memory is configured to carry out with one or more video modules the transmission of video pixel data;
Video compressing module is coupled to described video memory by signal bus, and is configured to
Use quantizer the piece of brightness and colourity YUV inputting video data to be compressed into the compressed video data of the every pixel bit number with minimizing, each piece comprises for N pixel of brightness with for M pixel of each colourity,
Wherein the piece compression is in the situation that do not utilize each piece pixel data in addition to carry out;
With described compressed Video Data Storage in described video memory; And
Video decompression module is configured to take out the piece of compressed video data and be have same format and approximately received and the video data through decompress(ion) of the original video data of compression by described video compressing module with the compressed video data decompress(ion) of storing in described video memory by any order;
Described video decompression module is configured to export described video data through decompress(ion).
2. device as claimed in claim 1, wherein said compression are that the total draught R according to the degree of controll block compression carries out.
3. device as claimed in claim 2, wherein said total draught R can represent by the mode of ratio metering or according to the bit number that comprises in resulting.
4. device as claimed in claim 2, also being included in when keeping total draught R is that brightness and colourity are selected identical or different compression level.
5. device as claimed in claim 4, unless wherein colourity has high levels of complexity, reach bottom line, make the bit of brightness use reach maximum otherwise the degree of described compression is selected to the bit that colourity is used.
6. device as claimed in claim 1, also be included in the Texture complication that compression blocks is estimated in described compression before.
7. device as claimed in claim 6, the bit number that wherein distributes for brightness and colourity in compressed piece is in response to estimates what Texture complication was determined.
8. device as claimed in claim 1, also comprise with filling bit and come the video data of wadding warp compression to keep the fixed size of compression blocks.
9. device as claimed in claim 1 wherein adopts from the information to the colourity compression of this piece the luminance compression of given.
10. device as claimed in claim 1 also comprises:
Carry out pixel prediction between compression period, described pixel prediction starts from the initial reference pixel of selecting in the central authorities of piece and defines right prediction direction and the left prediction direction that can process side by side on demand;
Wherein said reference pixel both keeps identical to right prediction direction and left prediction direction.
11. device as claimed in claim 10, wherein said pixel value prediction is predicted current pixel value based on previous pixel value.
12. device as claimed in claim 10, wherein during described pixel value prediction, the prediction of both direction is independently, carry out side by side thus right and left to prediction processing and required processing time will reduce half.
13. device as claimed in claim 1 also comprises according to the sub-block of the configuration of expecting and cutting apart.
14. device as claimed in claim 11 also comprises at least a portion of a plurality of possibility sub-block configuration is calculated the sub-block value at cost and abandoned the sub-block configuration that cost surpasses threshold value or available bit number.
15. device as claimed in claim 1, wherein said inputting video data receives from imageing sensor.
16. device as claimed in claim 1, wherein said device is integrated in video camera or still camera.
17. device as claimed in claim 1, wherein said inputting video data comprise have Y, Cr and the brightness of Cb form and the form of colourity YUV information.
18. device as claimed in claim 1, the bus bandwidth when wherein said compression and decompress(ion) are performed to reduce coding video frequency data and video memory demand.
19. device as claimed in claim 1, wherein said device were configured to before Video coding compression and decompress(ion) video data, sent compressed video data by network before decompress(ion) or before decompress(ion) with Video Data Storage in medium apparatus.
20. device as claimed in claim 1, wherein:
Described video compression utilizes nonlinear quantization to carry out; And
Adopt the quantization step of different accuracy when carrying out described nonlinear quantization.
21. a video encoder apparatus that is used for coding yuv video data comprises:
Dispose the video memory of signal bus, be configured to carry out with one or more video modules the transmission of video pixel data;
Video compressing module is coupled to described video memory by described signal bus, and is configured to
In the situation that do not adopt pixel data from the pixel beyond each piece, use quantizer brightness and colourity YUV inputting video data to be compressed into the compressed video data of the every pixel bit number with minimizing, each piece comprises for N pixel of brightness with for M pixel of each colourity;
With described compressed Video Data Storage in described video memory; And
Video decompression module is configured to take out the compressed video data stored in described video memory and is have same format and approximately received and the video data through decompress(ion) of the original video data of compression by described video compressing module with its decompress(ion); And
Coding module is configured to non-which block of video data of sequentially selecting and is taken out by described video decompression module and the coding video data through decompress(ion) to exporting from described decompression module.
22. the method for a compression and decompress(ion) yuv video data comprises:
Use quantizer by compression ratio R, each piece that inputting video data is compressed into for video data to be had the brightness of minimizing and/or the compressed block of video data of chrominance bit number, each piece comprises for N pixel of brightness with for M pixel of each colourity;
The compression of wherein said inputting video data is in the situation that be not used to carry out from the data of the pixel of compressed outside;
With described compressed Video Data Storage in video memory; And
For any block of video data that the order by any expectation takes out from described video memory, described compressed video data decompress(ion) is exported to generate through the video data of decompress(ion).
23. method as claimed in claim 22 also comprises:
Sequentially select block of video data and receive described video data through the decompress(ion) described video data of encoding from described memory in response to non-.
24. the method for a compression and decompress(ion) yuv video data comprises:
Use quantizer by compression ratio R, each piece that inputting video data is compressed into for video data to be had the brightness of minimizing and/or the compressed block of video data of chrominance bit number, each piece comprises for N pixel of brightness with for M pixel of each colourity;
Described compression is in the situation that be not used to from the data of the pixel of compressed outside, video data be carried out;
Each sub-block that will be performed in given of described compression is selected linearity or nonlinear quantization;
With described compressed Video Data Storage in video memory; And
For any block of video data that the order by any expectation takes out from described video memory, described compressed video data decompress(ion) is exported to generate through the video data of decompress(ion).
25. the method for a compression and decompress(ion) yuv video data comprises:
Use quantizer by compression ratio R, each piece that inputting video data is compressed into for video data to be had the brightness of minimizing and/or the compressed block of video data of chrominance bit number, each piece comprises for N pixel of brightness with for M pixel of each colourity;
Described compression is in the situation that be not used to from the pixel data of compressed outside, block of video data be carried out;
Estimated brightness Texture complication and chrominance texture complexity;
Be that brightness and colourity are selected compression level separately when keeping total draught R;
In response to the piece characteristic, each sub-block that be performed in given of described compression is selected linearity or nonlinear quantization;
Wherein between described compression period, the luminance compression processing and utilizing is from the information that the colourity compression of same block is processed;
With described compressed Video Data Storage in video memory;
To any moment after the described storage of described compressed video data, take out the piece of described video data by the order of any expectation from described video memory; And
For the piece that takes out, described compressed video data decompress(ion) is exported to generate through the video data of decompress(ion).
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US1517407P | 2007-12-19 | 2007-12-19 | |
| US61/015,174 | 2007-12-19 | ||
| US12/103,419 | 2008-04-15 | ||
| US12/103,419 US8401071B2 (en) | 2007-12-19 | 2008-04-15 | Virtually lossless video data compression |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101472182A CN101472182A (en) | 2009-07-01 |
| CN101472182B true CN101472182B (en) | 2013-06-19 |
Family
ID=40788592
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008101861727A Expired - Fee Related CN101472182B (en) | 2007-12-19 | 2008-12-19 | Virtually lossless video data compression |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101472182B (en) |
| CA (1) | CA2644379C (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102469306A (en) * | 2010-10-29 | 2012-05-23 | 华晶科技股份有限公司 | Image compression method |
| CN102868869B (en) * | 2011-07-06 | 2015-05-13 | 华晶科技股份有限公司 | Image compression method, decompression method and electronic device with fixed compression ratio |
| CN103096049A (en) * | 2011-11-02 | 2013-05-08 | 华为技术有限公司 | Video processing method and system and associated equipment |
| WO2013179101A1 (en) * | 2012-06-01 | 2013-12-05 | Freescale Semiconductor, Inc. | Method and apparatus for performing compression of image data |
| CN109672923B (en) * | 2018-12-17 | 2021-07-02 | 龙迅半导体（合肥）股份有限公司 | Data processing method and device |
| CN114501024B (en) * | 2022-04-02 | 2022-07-19 | 苏州浪潮智能科技有限公司 | Video compression system, method, computer readable storage medium and server |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1249836A (en) * | 1997-11-04 | 2000-04-05 | 皇家菲利浦电子有限公司 | Method and apparatus for processing DVD video |
| CN1273003A (en) * | 1997-08-29 | 2000-11-08 | 罗技公司 | Video capture and compression system and method for composite video |
| CN1628463A (en) * | 2001-02-12 | 2005-06-15 | 株式会社摩迩迪 | Appts. and method of coding moving picture |
-
2008
- 2008-11-21 CA CA 2644379 patent/CA2644379C/en not_active Expired - Fee Related
- 2008-12-19 CN CN2008101861727A patent/CN101472182B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1273003A (en) * | 1997-08-29 | 2000-11-08 | 罗技公司 | Video capture and compression system and method for composite video |
| CN1249836A (en) * | 1997-11-04 | 2000-04-05 | 皇家菲利浦电子有限公司 | Method and apparatus for processing DVD video |
| CN1628463A (en) * | 2001-02-12 | 2005-06-15 | 株式会社摩迩迪 | Appts. and method of coding moving picture |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101472182A (en) | 2009-07-01 |
| CA2644379A1 (en) | 2009-06-19 |
| CA2644379C (en) | 2015-01-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| RU2504107C2 (en) | Compressing video data without visible losses | |
| EP1995971B1 (en) | A method and device for realizing quantization in coding-decoding | |
| CN1726709B (en) | Method and device for encoding image of uncompressed digital video frequency sequence | |
| CN101472182B (en) | Virtually lossless video data compression | |
| US7218674B2 (en) | Method and device for controlling quantization scales of a video encoding bit stream | |
| US20130083855A1 (en) | Adaptive color space selection for high quality video compression | |
| JP6141295B2 (en) | Perceptually lossless and perceptually enhanced image compression system and method | |
| CN108574841B (en) | Coding method and device based on self-adaptive quantization parameter | |
| CN101507277A (en) | Image encoding/decoding method and apparatus | |
| JP2007503784A (en) | Hybrid video compression method | |
| CN1157080A (en) | Device and method for coding video pictures | |
| EP2360926A1 (en) | Image encoder and image decoder | |
| EP3363200B1 (en) | Data compression | |
| WO2008047865A1 (en) | Image quality evaluation method, image quality evaluation system and image quality evaluation program | |
| CN110419222A (en) | Method, apparatus and system for being coded and decoded to video data | |
| CN104639946A (en) | Apparatus and method for processing image | |
| US8351723B2 (en) | Methods and apparatus for generating and using visual error weights | |
| JPH09200758A (en) | Image encoder | |
| US6701019B1 (en) | Determining visually noticeable differences between two images | |
| US8428381B2 (en) | Image compression method with variable quantization parameter | |
| CN101406034B (en) | Compression scheme using qualifier watermarking and apparatus using the compression scheme for temporarily storing image data in a frame memory | |
| Kim et al. | Implementation of DWT-based adaptive mode selection for LCD overdrive | |
| US11515961B2 (en) | Encoding data arrays | |
| US20070253629A1 (en) | Image Processing Device and Image Forming Device Provided therewith | |
| CN114760475A (en) | Video encoding method and device, storage medium and electronic equipment |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130619 Termination date: 20151219 |
|
| EXPY | Termination of patent right or utility model |