CN1610408B - A method of dynamic regulating video-communication coding and related apparatus - Google Patents
A method of dynamic regulating video-communication coding and related apparatus Download PDFInfo
- Publication number
- CN1610408B CN1610408B CN 200310102468 CN200310102468A CN1610408B CN 1610408 B CN1610408 B CN 1610408B CN 200310102468 CN200310102468 CN 200310102468 CN 200310102468 A CN200310102468 A CN 200310102468A CN 1610408 B CN1610408 B CN 1610408B
- Authority
- CN
- China
- Prior art keywords
- coding
- picture
- block
- kenel
- video
- 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 - Lifetime
Links
- 238000000034 method Methods 0.000 title claims abstract description 87
- 230000001105 regulatory effect Effects 0.000 title 1
- 230000008569 process Effects 0.000 claims description 12
- 238000006073 displacement reaction Methods 0.000 description 102
- 238000012360 testing method Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 6
- 230000002457 bidirectional effect Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000011144 upstream manufacturing Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000013139 quantization Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 206010021703 Indifference Diseases 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000000875 corresponding effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Images
Abstract
A method for coding a video data into a video bit stream for a video coding device is provided. The video data includes at least one picture which includes a plurality of blocks, a coding and decoding system including the video coding device. A plurality of devices in the coding and decoding system share the same memory. The method including: receiving an available memory frequency width value permitted by the coding and decoding system at present; comparing the memory frequency width value with a first critical value and a second critical value to determine the coding type of one block in the plurality of blocks; and coding the block in the coding form; wherein if the available memory frequency width value is not less than the first critical value, the available frequency width value and the second critical value are compared.
Description
Technical field
The present invention relates to a kind of method of video coding, particularly relate to a kind ofly, dynamically adjust the method for the coding kenel of a coded bit stream in order to when memory band width that system allowed changes.
Background technology
Along with the lifting of electronic circuit arithmetic speed, the computing (handling as video signal) that expends a large amount of system resources becomes one of development priority of electronic installation then.Usually include some coding kenels in the specification that general video signal is handled, to adapt to the coding demand of various video signal datas.(Moving Picture Coding Experts Group 2, MPEG 2) is example with Motion Picture Experts Group's two type specifications, and it mainly is to eliminate on the video pictures space and temporal redundant information, to reach the video signal compression effects.
In general, having some common characteristics on same video pictures perhaps, is on the color, perhaps be on how much, or the further feature value obtains.Remove the mode of the redundant information on these spaces, will identify element important in the picture exactly, and remove repetition and do not have the element that influences.For example, in Motion Picture Experts Group's two type specifications (MPEG 2), be the mode of utilizing color sampling, discrete cosine transform and quantification, realize removing the purpose of the redundant information on the aforesaid space.In addition, the general video signal of playing is a succession of continuous images sequence in fact, yet because have a kind of existing picture of the persistence of vision on the human vision, so can produce the illusion of the image that links up.And in this kind video signal, because the time interval between picture is very little, so several indifferences between the picture that faces mutually are the change in location of picture material mostly.Therefore, in Motion Picture Experts Group's two type specifications (MPEG 2),, realize compressing the purpose of video signal also in the mode of the redundant information that similitude was caused between picture and picture on the elimination time shaft.
In order to remove aforementioned temporal redundant information, in Motion Picture Experts Group's two type specifications (MPEG 2), bit shift compensation (Motion Compensation, video signal compress technique MC) have been used.Before carrying out bit shift compensation, at first a video pictures is divided into many 16X16 pixels macro zone block (Macro-block, MB).Then, with a macro zone block is that unit is as a target macro block, and in a reference picture (Reference Picture), look for this target block approximate one with reference to macro zone block, and this target macro block to this coordinate difference with reference to position between macro zone block is recorded as a motion vector (Motion Vector).If this before being positioned at the residing picture of this target macro block on the time shaft, is called prediction (Forward Prediction) forward with reference to the residing picture of macro zone block; If this is positioned at after the residing picture of this target macro block with reference to the residing picture of macro zone block, be called prediction (Backward Prediction) backward on time shaft; If this target macro block with reference on the time shaft before it with the reference macro zone block of two direction reference pictures afterwards, then be called bi-directional predicted (Bi-directionallyPrediction).Above-mentioned process of looking for this motion vector is referred to as displacement prediction (MotionEstimation), and block matching method (Block Matching Method) is one of common displacement prediction mode.In addition, because this target macro block and this is with reference to not necessarily in full accord between macro zone block, therefore also to calculate this target macro block and this with reference to the picture signal difference between macro zone block, also be referred to as predicated error (Prediction Error), do the usefulness of compensation during in order to Zai Xie Code.
In Motion Picture Experts Group's two type specifications (MPEG 2), defined interior coding (Intra), predictive coding (Predictive), with bi-directional predictive coding three kinds of picture compressing and coding patterns such as (Bi-directionally Predictive).Wherein, an interior coded picture (Intra Picture), or be called the I picture, and be a picture of absolute coding, need not can carry out the coding of coded picture in this with reference to the picture of front and back.One prediction encoded picture (Predictive Picture), or be called the P picture needs the nearest reference picture (reference picture must be a coded picture or a prediction encoded picture in) in front on the reference time axle to carry out forward predictive coding during coding.In addition, a two-way prediction encoded picture (Bi-directionally Predictive Picture), or be called the B picture need be referenced to the data of the reference picture of front and back both direction during coding, forward encodes with backward-predicted carrying out.
Learn that according to above kind class declaration an interior coded picture is because of the operation that does not need can encode with reference to the picture of front and back, so its employed memory band width minimum in cataloged procedure about image frame.And a prediction encoded picture is because need with reference to reference picture in the past carrying out encoding operation, so its employed memory band width in cataloged procedure is bigger.By that analogy, a two-way prediction encoded picture because need with reference in the past and the data of following both direction reference picture carrying out encoding operation, so its employed memory band width maximum in cataloged procedure.
As previously mentioned, a picture is made up of a plurality of macro zone blocks (Macroblock), and is to be that unit encodes with the macro zone block when encoding this picture.Each macro zone block in a picture is used for representing that all to a macro zone block kenel parameter (Motion Type) should be arranged this macro zone block is to carry out the bit shift compensation coding in which way.With Motion Picture Experts Group's two type specifications (MPEG 2) is example, and in the interior coded picture, the coding kenel of all macro zone blocks is interior coding.And in the prediction encoded picture, the coding kenel of described macro zone block can be interior coding or forward shift compensation coding (Forward Motion Compensation).In addition, among the one two-way prediction encoded picture, the coding kenel of described macro zone block can be interior coding, forward shift compensation coding, upstream shift compensation coding (Backward Motion Compensation) or bidirectional displacement compensation coding (Bi-directional Motion Compensation).Wherein an interior coding macro zone block is a macro zone block of absolute coding, need not can carry out encoding operation with reference to the macro zone block of front and back picture.And forward shift compensation macro zone block must be in the reference picture in past a most close reference macro zone block, read one forward prediction data to encode.In addition, a two-way bit shift compensation macro zone block must be from the reference macro zone block of past and following reference picture, read forward with the backward-predicted data to encode.
Learn that according to above coding kenel an interior coding macro zone block is because of the operation that does not need can encode with reference to the macro zone block of front and back reference picture, so its employed memory band width minimum in the process of coding about macro zone block.And forward shift compensation macro zone block, because need the operation of macro zone block to encode of reference reference picture in the past, so its employed memory band width in cataloged procedure is bigger.In like manner, because a two-way bit shift compensation macro zone block need be with reference to macro zone block the operation to encode of past with the reference picture in future, so its employed memory band width maximum in cataloged procedure.
Yet, employed memory band width size in the macro zone block cataloged procedure, not only relevant, also relevant with its employed displacement prediction mode with its coding kenel.
Fig. 1 is for carrying out the schematic diagram of the displacement prediction of a macro zone block with the known zone BMA.As shown in Figure 1, a target picture 120 is distinguished into a plurality of blocks, and each block can be any size, yet in Motion Picture Experts Group's specification (MPEG), target picture 120 is to be distinguished into the macro zone block that a plurality of sizes are the 16X16 pixel.Each macro zone block in the middle of the target picture 120, be according to its with time shaft on difference between a macro zone block in the middle of the last picture 110, or and the central macro zone block of back one picture 130 between difference encode.When carrying out the block coupling with a target macro block 100, can compare one by one to the macro zone block to be compared of all similar sizes in the search area 115 of last picture 110, or to a search area 135 of back one picture 130 in the macro zone block to be compared of all similar sizes compare one by one.In Fig. 1, after the comparison through the block matching method, in the middle of all macro zone blocks to be compared of last picture and back one picture, find out is a macro zone block 150 with target macro block 100 difference minimums, is called with reference to macro zone block.With reference to the motion vector and the signal difference of 100 of macro zone block 150 and target macro blocks, just can be used as the foundation of coding.Therefore, when decompressing, just can utilize the coding of each macro zone block in the last picture 110, cooperate the motion vector and the signal difference of each macro zone block in the target picture 120, so that target picture 120 decodings are reduced.
With a kind of block matching algorithm (Block-Matching Algorithm) that all macro zone blocks to be compared in a target macro block and the search area are compared one by one, be referred to as the full search type block of single-order layer matching algorithm (Full Search Block-Matching Algorithm).Generally speaking, the big more motion vector more accurately of trying to achieve of search area, however cost is that employed memory band width also can be directly proportional with the area of search area in the comparison process.For example, when using the full search type block of single-order layer matching algorithm to seek a size as the reference macro zone block of the macro zone block of 16X16 pixel, if search area is a positive and negative N pixel, and accuracy is a pixel, then needs to carry out (2N+1)
2Inferior block comparison operation.In other words, if N is 16, then representative need be carried out the block comparison operation of 1089 16X16 sizes.Because the comparison of block each time all needs to carry out 256 times calculating, so this kind algorithm can consume a large amount of computing energy and memory band width.
Figure 2 shows that the schematic diagram of a known coding/decoding system 10, it includes a video information coder 12, a video decoder 14, a message coder 16, a central control unit 18, a memory management circuitry 20, a display control circuit 22, reaches a storage arrangement 24.As shown in Figure 2, in known coding/decoding system 10, all devices are shared same memory management circuitry 20 and same storage arrangement 24, and wherein memory management circuitry 20 is to be used for the operation of each device access memory device 24 in the management system 10.
Yet as previously mentioned, needed memory band width was very huge when video information coder 12 was encoded.When the frequency range of storage arrangement 24 has in limited time, in case video information coder 12 carries out the coding mode of consume memory frequency range maximum, just may cause needed memory band width to heighten suddenly and lower the usefulness of coding, so known coding/decoding system 10 can't be realized the requirement of real-time coding constantly.Thereby can coding/decoding system 10 realize the requirement of real-time coding, just becomes a crucial problem.
In order to overcome the shortcoming that the full search type method of single-order layer consumes too many memory band width, people such as H.M.Jong proposed a kind of three-layered formula search method (see " Parallel Architectures for 3-StepHierarchical Search Block-Matching Algorithm; " IEEE Trans.OnCircuits and Systems for Video Technology, Vol.4, August 1994, pp.407-416).Usually stratum's search type mode then, can be dwindled the area of search gradually carrying out a rough search earlier in a big way one at the beginning in following step.Under the identical situation of search area, the employed memory band width of hierarchy search method is really than the full search method of single-order layer employed lacking, and it is better to carry out efficient (Performance), but image quality is reduced.
People such as Maturi propose the notion of full search method of a kind of single-order layer and hierarchy search method applied in any combination in US Patent 5731850.Determine one in order to carry out the search area of block comparison according to the type of coding of a target picture, when this search area surpasses a default value, adopt the hierarchy search method to carry out the block comparison, when if search area is less than or equal to this default value, then adopt the full search method of single-order layer to carry out the block comparison, in the hope of on image quality and execution efficient, obtaining better balance.The method that people such as Maturi propose is the mode that changes the block comparison according to the size of search area.Yet, under the system memory bandwidth condition of limited, once the needed memory band width of video information coder is when heightening suddenly, people's such as Maturi method also can't be guaranteed the employed memory band width of video information coder, can not surpass the maximum memory frequency range that system can provide, therefore, system storage can can't be loaded and lowered the usefulness of encoding, and can't realize the requirement of real-time coding.
Summary of the invention
In view of this, the invention provides video encoding method and relevant apparatus, in order to the coded system of dynamic adjustment one video signal data, so that employed memory band width can not surpass the available memory frequency range that system provided in the cataloged procedure.
The present invention has disclosed a kind of method that a video coding device is encoded into video signal data one video signal bit stream that is used in embodiment, wherein, this video signal data comprises at least one picture, one coding/decoding system comprises this video coding device, multiple arrangement in this coding/decoding system is shared a memory, and this method includes: receive the current available memory frequency range value that allows of this coding/decoding system; Relatively this available memory frequency range value and first critical value and second critical value are to determine a coding kenel of this picture; And this picture is encoded with this coding kenel; Wherein be not less than first critical value, then compare the available memory frequency range value and second critical value as available memory frequency range value.
The present invention has also disclosed a kind of embodiment of video coding device, be used for video signal data is encoded into a video signal bit stream, one coding/decoding system comprises this video coding device, multiple arrangement in this coding/decoding system is shared a memory, and this video coding device includes: means are used for receiving the current available memory frequency range value that allows of this coding/decoding system; One means are used for relatively this available memory frequency range value and first critical value and second critical value to determine a coding kenel of this video signal bit stream; And one means be used for this video signal data being encoded with this coding kenel; Wherein be not less than first critical value, then compare the available memory frequency range value and second critical value as available memory frequency range value.
The advantage of video encoding method of the present invention and code device is that employed memory band width can not surpass the maximum memory frequency range that system provided in the cataloged procedure, and can realizes the requirement of real-time coding.
Description of drawings
Fig. 1 carries out the schematic diagram of the displacement prediction of macro zone block for known technology;
Fig. 2 is the schematic diagram of a known coding/decoding system;
Fig. 3 is a video coding device of the present invention;
Fig. 4 is the flow chart of video encoding method of the present invention; With
Fig. 5 is the flow chart of another embodiment of video encoding method of the present invention.
The drawing reference numeral explanation
10 coding/decoding systems, 100,150 macro zone blocks
12 video information coders, 110,120,130 pictures
14 video decoders, 115,125,135 search areas
16 message coders, 200 video coding devices
18 central control units, 202 video signal interfaces
20 memory management circuitry, 204 displacement prediction devices
22 display control circuits, 206 coding circuits
24 storage arrangements
Embodiment
Please refer to Fig. 3.Fig. 3 is the schematic diagram of video coding device 200 of the present invention.Video coding device 200 is in order to digital video data (Digitized Video Data) are encoded, to produce a coded bit stream.Video coding device 200 includes a video signal interface (Video Interface) 202, in order to receive digital video data; One displacement prediction device 204 is electrically connected in video signal interface 202, in order to the macro zone block in these digital video data is carried out displacement prediction; An and coding circuit 206, be electrically connected in video signal interface 202 and displacement prediction device 204, in order to control whole coding flow process and the digital video data of this process displacement prediction encoded, for example carry out discrete cosine transform (Discrete CosineTransform), quantization operations (Quantization), with variable length code computings such as (Variable LengthCoding), to produce a coded bit stream and to carry out inverse quantization computing (InverseQuantization), reverse discrete cosine conversion (Inverse Discrete Cosine Transform) and bit shift compensation (Motion Compensation) are to produce the needed reference picture of displacement prediction.Wherein video coding device 200 is electrically connected in a memorizer control circuit (not shown) as shown in Figure 2, with access one system storage (not shown).
After the video signal interface in the video coding device 200 202 receives digital video data, can carry out operating to these digital video data such as pre-process such as noise removing, picture convergent-divergents.Then, by the digital video data that video signal interface 202 will be handled, be sent to coding circuit 206 and displacement prediction device 204.At this moment, displacement prediction device 204 is used the video encoding method of a kind of dynamic adjustment proposed by the invention with coding circuit 206, and the information of this memorizer control circuit receiving system memory band width certainly is to adjust the coded system of this coded bit stream.Below will describe detailed operation mode of the present invention in detail.
Please refer to Fig. 4.Fig. 4 is the flow chart 300 of the video encoding method of the disclosed a kind of dynamic adjustment of the present invention, and the step of this method is as follows:
Step 302: beginning.
Step 304: coding circuit 206 receives a frequency range information of this system storage.
Step 306: coding circuit 206 judges according to received frequency range information whether frequency range enough becomes a picture coding one prediction encoded picture or a two-way prediction encoded picture.If the frequency range deficiency, then decision is a coded picture in this picture coding, and carry out step 308; If frequency range is enough, then according to received frequency range information, decision is a prediction encoded picture or a two-way prediction encoded picture with this picture coding, and carry out step 311.
Step 308: coding circuit 206 is encoded to the block in this picture with interior coding kenel.
Step 310: check whether this block is block last in this picture, if not, then carry out step 308 once more, and next block is encoded; If this block has been block last in this picture, then carry out step 326.
Step 311: displacement prediction device 204 receives a frequency range information of this system storage.
Step 312: displacement prediction device 204 judges according to received frequency range information whether frequency range enough carries out displacement prediction to the block in this picture, if the frequency range deficiency then carry out step 322; If frequency range is enough, then carry out step 314.
Step 314: coding circuit 206 carries out encoded test with interior coding kenel to this block.
Step 316: displacement prediction device 204 determines at least one displacement prediction kind according to the frequency range situation, and adjusts corresponding search area.Displacement prediction device 204 also carries out displacement prediction with this at least one displacement prediction kind to this block respectively, with predicated error (Prediction Error), determine one to have the best displacement prediction mode of minimum predicated error by more different displacement prediction modes.
Step 318: coding circuit 206 carries out the result of displacement prediction with this best displacement prediction mode to this block according to displacement prediction device 204, with displacement prediction coding kenel this block is carried out encoded test.
Step 320: the encode test results of two kinds of coding kenels in comparison step 314 and the step 318, and therefrom select a kind of coding kenel, according to this this block is encoded.
Step 322: this block coding is become encoded block in interior coding kenel.
Step 324: check whether this block is block last in this picture, if not, then carry out step 312 once more, and next block is encoded; If this block has been block last in this picture, then carry out step 326.
Step 326: because this block has been the last block in this picture, so finish the coding of this picture.
Below with the operation situation of a preferred embodiment description of flow diagram 300 of the present invention among Fig. 3.In step 302, the digital video data that the video signal interface 202 of Fig. 3 will be handled are sent to displacement prediction device 204 and coding circuit 206.Then in step 304, video coding device 200 receives a system memory bandwidth information from this memorizer control circuit, and this memory band width packets of information contains the maximum memory frequency range that video coding device 200 is allowed to use.And displacement prediction device 204 of the present invention promptly is according to this system memory bandwidth size with coding circuit 206, dynamically adjusts the mode of running.
In step 306, coding circuit 206 is judged the coding kenel of the picture (a frame picture (Frame Picture) or a picture (FieldPicture)) that promptly will encode in the coded bit stream according to the memory band width that is allowed.If system allowed the memory band width that video coding device 200 uses at that time, only enough support this picture to be encoded, then in step 306, just coding circuit 206 can be selected with interior coding kenel as the coded system to this picture with interior coding kenel.Then, in step 308 to step 310, because coding circuit 206 decision is encoded to this picture with interior coding kenel, therefore, for each block in the middle of this picture, coding circuit 206 all can be encoded with interior coding kenel, and each block all is encoded into interior encoded block.In Motion Picture Experts Group's specification (MPEG) series, above-mentioned block is to be unit with the macro zone block.
On the contrary, in step 306, if coding circuit 206 is judged the memory band width that this moment, system allowed, enough support this picture to be encoded with predictive coding kenel or bi-directional predictive coding kenel, then according to the memory band width size of this permission, the coding kenel that determines this picture is predictive coding or bi-directional predictive coding.For example, in coding circuit 206, store one and represent required minimized memory frequency range critical value (Threshold) Pth of predictive coding kenel, and one represent the required minimized memory frequency range critical value Bth of bi-directional predictive coding kenel.When this moment memory band width that system is allowed more than or equal to this memory band width critical value Bth, then this picture is encoded with the bi-directional predictive coding kenel, be a two-way prediction encoded picture with this picture coding; Otherwise, when the memory band width that allows less than this memory band width critical value Bth, but, then this picture is encoded with the predictive coding kenel more than or equal to this memory band width critical value Pth, be a prediction encoded picture with this picture coding.
In one embodiment of this invention, this memory band width critical value Pth and Bth can be set at the estimated value that is predetermined.In another embodiment of the present invention, coding circuit 206 is in the process of coding, also can add up employed memory band width in the cataloged procedure of a plurality of pictures simultaneously, and employed minimized memory frequency range value when up to the present carrying out predictive coding respectively, and employed minimized memory frequency range value when carrying out bi-directional predictive coding, upgrade this memory band width critical value Pth and Bth, with the foundation when judging next picture coding kenel.
Next in step 311, the information of displacement prediction device 204 receiving system memory band widths.Please note, in step 304, the system memory bandwidth information that coding circuit 206 is received, be to correspond to the memory band width information than long duration, and displacement prediction device 204 received system memory bandwidth information in step 311, be the memory band width information that corresponds in the short period, make video coding device 200 of the present invention do the most effective utilization to system storage according to the memory band width information of different time points.
Then, in step 312, displacement prediction device 204 is made comparisons with a critical value Mth according to received memory band width information, and this critical value Mth represents to desire to carry out the required minimized memory frequency range of displacement prediction.To judge whether enough this block to be carried out displacement prediction.If the memory band width that allow this moment is not enough to support this block is carried out the operation of displacement prediction, then in step 322, just coding circuit 206 becomes encoded block in this block coding.In one embodiment of this invention, carry out required this memory band width critical value Mth of displacement prediction, can be set at an estimated value that is predetermined.In another embodiment of the present invention, displacement prediction device 204 is in the process of carrying out displacement prediction, also can add up employed memory band width in the displacement prediction process of a plurality of blocks simultaneously, and upgrade this critical value Mth according to up to the present carrying out the required memory band width minimum value of displacement prediction, with as the foundation that judges whether next block is carried out displacement prediction.
On the other hand, in step 312, if displacement prediction device 204 judges that the memory band width that was allowed at that time is enough to this block is carried out displacement prediction, then next in step 314, coding circuit 206 carries out encoded test with interior coding kenel to this block, for follow-up ratio than usefulness.
In step 316, displacement prediction device 204 is according to the frequency range situation that is allowed, and the coding kenel (predictive coding kenel or bi-directional predictive coding kenel) of this picture of in step 306, being determined of coding circuit 206, to select to correspond at least one displacement prediction kind of this block.Then, displacement prediction device 204 carries out displacement prediction with this at least one displacement prediction kind to this block respectively, and the predicated error that produced of more different displacement prediction mode.Displacement prediction device 204 therefrom determines one to have the best displacement prediction mode of minimum predicated error according to relatively result, and should the resulting motion vector of the best displacement prediction mode etc. information be sent to coding circuit 206.
As defined in Motion Picture Experts Group's specification (MPEG), when the coding kenel of this picture is the predictive coding kenel, coding kenel or forward shift compensation coding kenel in the coding kenel of this block can be.When the coding kenel of this picture is bi-directional predictive coding, the coding kenel of this block can be the kenel (but stipulate in Motion Picture Experts Group's specification four types (MPEG 4): when the coding kenel of this picture was bi-directional predictive coding, can not there be interior coding kenel in the coding kenel of central each block) of interior coding, forward shift compensation coding, upstream shift compensation coding or bidirectional displacement compensation coding.When carrying out the coding of this block with interior coding kenel, displacement prediction device 204 does not need to carry out the operation of displacement prediction.When carrying out the coding of this block with forward shift compensation coding kenel, displacement prediction device 204 need carry out the forward shift prediction.When carrying out the coding of this block with upstream shift compensation coding kenel, displacement prediction device 204 need carry out the upstream shift prediction.When carrying out the coding of this block with bidirectional displacement compensation coding kenel, displacement prediction device 204 need carry out the bidirectional displacement prediction.
In addition, with Motion Picture Experts Group's two type specifications (MPEG 2) is example, in step 316, displacement prediction device 204 is except determining the displacement prediction kind of this block, also need be according to the kind (a frame picture or a picture) of this picture and the coding kenel of this picture, to determine displacement kenel (MotionType) parameter of this block.Displacement kenel parameter is to make the bit shift compensation coding in which way in order to indicate this block.For example, when this picture is a frame picture, and when its coding kenel is bi-directional predictive coding, if 204 decisions of displacement prediction device are carried out the forward shift prediction to this block, then displacement prediction device 204 also need select the displacement kenel parameter of this block be frame be the base prediction kenel (Frame-Based Prediction), or be the prediction kenel (Field-Based Prediction) of base, to carry out the foundation of bit shift compensation when encoding as 206 pairs of these blocks of coding circuit.
In other words, in step 316, displacement prediction device 204 is according to memory band width situation and correlated condition, determines at least a displacement prediction kind and the displacement kenel that corresponds to this block.For example, suppose memory band width information received in step 311, to allow displacement prediction device 204 be base (Frame-Based) with the frame with this block is carried out the forward shift prediction for base (Field-Based) two kinds of displacement kenels, then displacement prediction device 204 can carry out the forward shift prediction with above-mentioned two kinds of displacement kenels to this block respectively, then, the predicated error (PredictionError) that compares two kinds of forward shift prediction mode.Displacement prediction device 204 is therefrom selected a best displacement prediction mode with minimum predicated error, and should the resulting motion vector of the best displacement prediction mode etc. information be sent to coding circuit 206.
As previously mentioned, in step 316, displacement prediction device 204 carries out the displacement prediction of this block normally with single-order layer full search type block matching method or hierarchy search block matching method.In the middle of the preferred embodiment of the disclosed video encoding method of the present invention, an important techniques is characterised in that displacement prediction device 204 is according to the memory band width that is allowed, and adjusts the search area of block matching method.When displacement prediction device 204 was adopted the full search type block of single-order layer matching method this block is carried out displacement prediction, displacement prediction device 204 according to the memory band width that is allowed, was adjusted the search area of the full search type block of single-order layer matching method in the middle of step 316.When the memory band width that allows is big more, the search area that displacement prediction device 204 sets also can be big more.Otherwise when the memory band width that allows is more little, the search area that displacement prediction device 204 sets also can be more little.In like manner, when displacement prediction device 204 adopted stratum's search type block matching method, displacement prediction device 204 can be adjusted the search stratum number of stratum's search type block matching method and the search area in each stratum according to the memory band width that is allowed in step 316.In another embodiment of the present invention, displacement prediction device 204 can comprise single-order layer full search type block matching method and hierarchy search block matching method simultaneously, and, select to use single-order layer full search type block matching method or hierarchy search block matching method according to the memory band width that is allowed.For example,, can adopt hierarchy search block matching method when the memory band width that is allowed during less than a critical value Fth, otherwise, the full search type block of single-order layer matching method can be adopted.When selecting the full search type block of single-order layer matching method, displacement prediction device 204 of the present invention can be adjusted the search area of the full search type block of single-order layer matching method simultaneously accordingly; When selecting hierarchy search block matching method, displacement prediction device 204 of the present invention also can be adjusted the search stratum number of stratum's search type block matching method and the search area in each stratum simultaneously accordingly.
In step 318, coding circuit 206 carries out the information such as motion vector that this best displacement forecasting institute produces according to 204 pairs of these blocks of displacement prediction device, with displacement prediction coding kenel this block is carried out encoded test.
According to method provided by the invention, video coding device 200 is before each block (can be a macro zone block at this this block) coding, can carry out the encoded test of various different coding kenels earlier, to find out the coding kenel of using minimum bits of coded, with actual coding kenel as this block.Therefore, in the middle of step 320, the encode test results of the resulting displacement prediction coding of coding circuit 206 meeting steps 318 kenel, compare with the encode test results of the resulting interior coding kenel of step 314, consider to select in the middle of two kinds of coding kenels according to the design of coding circuit 206 better suited a kind of, according to this as the actual coding kenel of this block.
Because the mean value of each pixel (Pixel) brightness or the size of amount of variability statistics such as (Variance) in this block, it is the degree of this block image complexity of reaction, also react this block and save the coding kenel of storage volume most, that is the coding kenel of the minimum bits of coded of a use.Therefore, in step 320, coding circuit 206 is when comparing the encode test results of two kinds of coding kenels, can compare statistics such as the mean value of each pixel intensity in this block that encoded test in the step 314 calculates gained or amount of variability, calculate the size that this block carries out statistics such as the mean value of the predicated error that this best displacement forecasting institute obtains or amount of variability with encoded test in the step 318, decide an optimal coding kenel.Normally with the coding kenel of using minimum bits of coded actual coding kenel as this block.But coding circuit 206 is except can selecting the actual coding attitude of the preferable coding kenel of video signal compression ratio as this block, also can select the actual coding kenel of the preferable coding kenel of video signal quality as this block, the foundation of selection is looked closely the demand of circuit design and consideration and is decided.
Note that from the explanation of step 320 and can find that step 314 is not limited to and will carries out before step 316.In fact, step 314 can be between step 316 and step 318, even also can carry out after step 318, does not all influence effect of the present invention.
Step 324 is similar to step 310, is all coding circuit 206 and judges whether this block is last central block of this coded picture.If this block has been last block of this picture, represent then that this picture has been encoded and finish, if last block that this block is not this picture, then presentation code circuit 206 will carry out the operation of step 312 to step 324 to next block in this picture with displacement prediction device 204.
Please note, because after this block is finished coding, the memory band width that system allows or may change, so when coding circuit 206 is encoded with next block of 204 pairs of these coded pictures of displacement prediction device, displacement prediction device 204 is in step 312, can study and judge according to memory band width information instantly once more, whether carry out displacement prediction with decision.
After all blocks in this picture were all finished coding one by one, before the encoding operation that carries out next picture, coding circuit 206 is the information of receiving system memory band width once more, will encode to next picture with which kind of kenel of encoding with decision.
Note that in the middle of the explanation of the preferred embodiment of the invention described above, displacement prediction device 204 employed block matching methods, can but be not limited to two kinds of single-order layer full search type block matching method and stratum's search type block matching methods.As long as in the middle of calculating process, the proportional any block matching algorithm of the memory band width of employed block matching and searching scope and consumption, video coding mode of the present invention all can be supported.
In addition, in the above embodiments explanations, be to be example with Motion Picture Experts Group's two type specifications (MPEG 2).In the specification of Motion Picture Experts Group's four type specifications (MPEG 4), regulation is when the coded system of a picture is bi-directional predictive coding, any block in the middle of this picture all not can be encoded into interior encoded block, therefore, video encoding method of the present invention also can be simplified to support can not have in the specific picture coding kenel regulation of interior encoded block.
Please refer to Fig. 5 (and in the lump with reference to figure 4).Figure 5 shows that another embodiment flow chart 400 of video encoding method of the present invention.Flow chart 400 is the results after the flow chart 300 among Fig. 4 is simplified.Its function mode is similar to flow chart 300 haply, and between the two the most significantly difference be, the specific picture coding kenel of encoded block in flow chart 400 is applicable to and can not contains, for example, regulation in Motion Picture Experts Group's four type specifications (MPEG 4): a two-way prediction encoded picture can not contain interior encoded block.As long as the flow chart 300 of Fig. 4 is added that suitable judgement formula just can be equal to the declaration of will of flow chart 400, only be to be placed on Fig. 5 separately for the convenience on illustrating, the notion of video encoding method in fact of the present invention is applicable to the picture and the block of any coding kenel herein.
Combine with described, the most important technical characterictic of video encoding method of the present invention is the starting point considered with being restricted to of memory band width, according to maximum frequency range adjustment that system allowed coding kenel and the displacement prediction mode to a picture or a block.When finishing the coding of a block, adjust the coded system of next block once more according to new memory band width situation, when making each block in the middle of this picture of coding, employed memory band width is all in the maximum memory frequency range scope that system allowed at that time.Die, use the video coding system or the device of video encoding method of the present invention, employed at any time memory band width can be above the memory band width of system's maximum.Therefore, the memory band width that uses because of coding circuit in the middle of the known technology heightens and may cause the situation of real-time coding operation disruption, all can be resolved under application of the present invention.
The above only is preferred embodiment of the present invention, and all equalizations of doing according to claim of the present invention change and modify, and all should belong to the covering scope of patent of the present invention.
Claims (9)
1. one kind is used for the method that a video coding device is encoded into video signal data one video signal bit stream, this video signal data comprises at least one picture, one coding/decoding system comprises this video coding device, and the multiple arrangement in this coding/decoding system is shared a memory, and this method includes:
(a) receive the current available memory frequency range value that allows of this coding/decoding system;
(b) relatively this available memory frequency range value and one first critical value and one second critical value with determine this picture one the coding kenel; And
(c) with this coding kenel this picture is encoded;
Wherein be not less than this first critical value, then relatively this available memory frequency range value and this second critical value as this available memory frequency range value.
2. the method for claim 1, wherein step (b) includes:
When this available memory frequency range value during less than this first critical value, the coding kenel of setting this picture is interior coding.
3. method as claimed in claim 2, wherein step (b) also includes:
When this available memory frequency range value was between this first critical value and this second critical value, the coding kenel of setting this picture was predictive coding.
4. method as claimed in claim 3, it also includes:
In the process of this video signal data of coding, upgrade this first or second critical value adaptively.
5. method as claimed in claim 2, wherein step (b) also includes:
When this available memory frequency range value during greater than this second critical value, the coding kenel of setting this picture is a bi-directional predictive coding.
6. method as claimed in claim 5, it also includes:
In the process of this video signal data of coding, upgrade this first or second critical value adaptively.
7. the method for claim 1, wherein this picture is a frame picture or a picture.
8. the method for claim 1, wherein this coding kenel is interior coding, predictive coding or bi-directional predictive coding.
9. a video coding device is used for video signal data is encoded into a video signal bit stream, and a coding/decoding system comprises this video coding device, and the multiple arrangement in this coding/decoding system is shared a memory, and this video coding device includes:
One means are used for receiving the current available memory frequency range value that allows of this coding/decoding system;
One means are used for relatively this available memory frequency range value and one first critical value and one second critical value to determine a coding kenel of this video signal bit stream; And
One means are used for this coding kenel this video signal data being encoded;
Wherein be not less than this first critical value, then relatively this available memory frequency range value and this second critical value as this available memory frequency range value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200310102468 CN1610408B (en) | 2003-10-21 | 2003-10-21 | A method of dynamic regulating video-communication coding and related apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200310102468 CN1610408B (en) | 2003-10-21 | 2003-10-21 | A method of dynamic regulating video-communication coding and related apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1610408A CN1610408A (en) | 2005-04-27 |
| CN1610408B true CN1610408B (en) | 2011-12-21 |
Family
ID=34756405
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200310102468 Expired - Lifetime CN1610408B (en) | 2003-10-21 | 2003-10-21 | A method of dynamic regulating video-communication coding and related apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1610408B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103974060A (en) * | 2013-01-31 | 2014-08-06 | 华为技术有限公司 | Method and device for adjusting video quality |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101340575B (en) * | 2007-07-03 | 2012-04-18 | 英华达(上海)电子有限公司 | Method and terminal for dynamically regulating video code |
| US9355613B2 (en) | 2012-10-09 | 2016-05-31 | Mediatek Inc. | Data processing apparatus for transmitting/receiving compression-related indication information via display interface and related data processing method |
| US10136160B2 (en) * | 2014-10-24 | 2018-11-20 | Mediatek Inc. | Method and apparatus for resolution adaptive video encoding with search range constraint |
-
2003
- 2003-10-21 CN CN 200310102468 patent/CN1610408B/en not_active Expired - Lifetime
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103974060A (en) * | 2013-01-31 | 2014-08-06 | 华为技术有限公司 | Method and device for adjusting video quality |
| CN103974060B (en) * | 2013-01-31 | 2016-01-27 | 华为技术有限公司 | Video quality method of adjustment and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1610408A (en) | 2005-04-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101459844B (en) | Selective and/or scalable complexity control for video codecs | |
| JP5318561B2 (en) | Content classification for multimedia processing | |
| US20100166073A1 (en) | Multiple-Candidate Motion Estimation With Advanced Spatial Filtering of Differential Motion Vectors | |
| KR970032142A (en) | Method and circuit for determining quantization interval in video encoding apparatus | |
| JPH07288810A (en) | Moving image coding method | |
| US10284850B2 (en) | Method and system to control bit rate in video encoding | |
| JP4101034B2 (en) | Encoding apparatus and method | |
| US7961787B2 (en) | Method of video content complexity estimation, scene change detection and video encoding | |
| KR20030009669A (en) | multi channel image encoding apparatus and encording method thereof | |
| CN101931800B (en) | Method and system for encoding region of interest by limited variable bit rate control | |
| CN1610408B (en) | A method of dynamic regulating video-communication coding and related apparatus | |
| US9137537B2 (en) | Dynamic reference frame decision method and system | |
| US20080205527A1 (en) | Video decoding device and method, and video coding device | |
| CN101998117B (en) | Video transcoding method and device | |
| JP2000059792A (en) | High efficiency encoding device of dynamic image signal | |
| CN103002284B (en) | A kind of video coding-decoding method based on model of place adaptive updates | |
| CN101977317B (en) | Intra-frame prediction method and device | |
| JP2914607B2 (en) | Image signal decoding device | |
| WO2007006181A1 (en) | A rate control method and apparatus | |
| JPH06169452A (en) | Picture compression system having weighting on screen | |
| KR20030009575A (en) | Rate control apparatus and method using neural-net classifier for moving picture coding | |
| Jubran et al. | Sequence-level reference frames in video coding | |
| JPH07111653A (en) | Encoder | |
| JPH11346365A (en) | Encoding/compression method and device | |
| US6678327B1 (en) | Image signal compression coding method and apparatus |
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 | ||
| CX01 | Expiry of patent term |
Granted publication date: 20111221 |
|
| CX01 | Expiry of patent term |