CN100393137C - Moving image coding apparatus - Google Patents
Moving image coding apparatus Download PDFInfo
- Publication number
- CN100393137C CN100393137C CNB2005100772084A CN200510077208A CN100393137C CN 100393137 C CN100393137 C CN 100393137C CN B2005100772084 A CNB2005100772084 A CN B2005100772084A CN 200510077208 A CN200510077208 A CN 200510077208A CN 100393137 C CN100393137 C CN 100393137C
- Authority
- CN
- China
- Prior art keywords
- intra prediction
- unit
- image
- prediction mode
- image block
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
An image coding apparatus determines an image pattern of image data and, based on the determined image pattern, selects a prediction mode for generating predicted pixel values by predicting pixel values in a frame using pixel values in the same frame. Alternatively, based on photographing information concerning input image data, an image coding apparatus selects a prediction mode for generating predicted pixel values by predicting pixel values in a frame using pixel values in the same frame.
Description
Technical field
The present invention relates to a kind of moving image coding apparatus, relate more specifically to a kind of technology that when using coding standard H.264, is used to select the optimal frames inner estimation mode.
Background technology
Set up various coding standards, for example Motion-JPEG, MPEG-1 and MPEG-2 are as the technology that is used for mobile image is carried out high efficient coding.Manufacturer has developed and the selling images trap setting, for example digital camera, Digital Video, DVD (digital versatile dish) player etc., and it can store the mobile image that has used these coding standards.Therefore, allow the user to use these image capture devices, DVD player, PC etc. easily mobile image to be play.
Digitlization is moved image and is carried lot of data.So the various mobile method for encoding images that can carry out than the above-mentioned more effective high compression of coding standard such as MPEG-1 or MPEG-2 are studied continuously and develop.Recently, a kind of new encryption algorithm of Part 10AVC (back is called H.264 standard) that is called H.264/MPEG-4 is by ITU-T (International Telecommunications Union-telecommunication standardization group) and ISO (International Standards Organization) standardization.
H.264 standard with compare such as traditional coding standard of MPEG-1 and MPEG-2, require the very big computation complexity of Code And Decode, but provided higher code efficiency.Disclosed the H.264 system and method for the computing of standard of a kind of use, for example, in Japanese Laid-Open Patent Application No.2004-56827.
H.264 standard comprises a kind of Forecasting Methodology that is called infra-frame prediction, and this Forecasting Methodology is predicted pixel value in a given frame by using the pixel value in the same frame.In this infra-frame prediction, many intra prediction modes of using selected are arranged.In this case, even still have the coded data of minimum deterioration after selecting a kind of intra prediction mode of suitable input picture efficiently to compress with formation.
For infra-frame prediction, H.264 standard provides nine kinds of intra prediction modes in order to improve accuracy of predicting.Usually by tentatively input picture being carried out all intra prediction modes and found a kind of intra prediction mode that can obtain optimal result according to experimental execution result, from nine kinds of intra prediction modes, select a kind of optimal frames inner estimation mode.
The reason that adopts this system of selection is described below with reference to Figure 17.Figure 17 is the figure of expression frame, and wherein Wai Bu quadrangle is represented whole image, and the quadrangle of five inside is represented to carry out the square of infra-frame prediction.Though in fact on whole image, all be provided with square, only illustrate five representational squares for convenience of description.Arrow among Figure 17 is represented by being confirmed as the represented prediction direction of optimal frames inner estimation mode as the result of calculation in each square.Like this because even exist all types of target on each position in the same painting face, be confirmed as optimum so have different intra prediction modes, and may select and surpass one concrete intra prediction mode at each square.Based on this reason, when carrying out infra-frame prediction, intra prediction modes all in each square is obtained accuracy of predicting, and select the intra prediction mode that to carry out optimum prediction and specify in each square.
Yet, calculating all intra prediction modes in each square so that therefrom select a kind of optimal frames inner estimation mode, this has increased the complexity of calculating in the encoding process H.264, has therefore caused the extra increase of encoding process time or the waste of electric power.
Summary of the invention
The present invention has considered above-mentioned situation, and one aspect of the present invention is to help the optimal frames inner estimation mode of selection for input picture.
Another aspect of the present invention is to help to select optimal frames inner estimation mode for input picture by the shooting information of using relevant input picture.
Thereby another aspect of the present invention is to realize efficient coding by the computation complexity that alleviates in the picture coding device that uses standard H.264.
In one aspect of the invention, picture coding device comprises: the input unit that is configured to input image data, thereby be configured to by being divided into the cutting unit that the square that comprises a plurality of pixels respectively generates piece by the view data of input unit input, the picture pattern of the view data in each piece that is configured to determine to be generated by cutting unit is order unit really, be mixed with according to one the selected cell of selecting by the definite picture pattern of determining unit in a plurality of predictive modes, and be configured to by using the pixel value in the same painting face to carry out the processing unit of picture pixel value prediction processing with the pixel value of the prediction of output according to the predictive mode of selecting by selected cell.
In another aspect of this invention, picture coding device comprises: be configured to import will the image encoded data input unit, be configured to obtain about the photographic information of view data acquiring unit, be configured to use described photographic information from n predictive mode, to select m predictive mode (thereby the selected cell of 1≤m≤n) and be configured to use pixel value in the same painting face the pixel value in this picture to be carried out the processing unit of the pixel value of the prediction processing prediction of output by m the predictive mode of selecting according to described selected cell.
By reading below in conjunction with the detailed description of accompanying drawing to relevant embodiment, other characteristic of the present invention and advantage will become obviously for those skilled in the art, and wherein same label is represented same or similar part in whole accompanying drawing.
Description of drawings
With the description of drawings embodiments of the invention specification one and that form a specification part, and explain principle of the present invention with this description.
Fig. 1 is the block diagram of explanation according to the intra-prediction process of the first embodiment of the present invention.
Fig. 2 is the schematic diagram of explanation Hadamard conversion.
Fig. 3 A, 3B and 3C are the schematic diagram of conversion example in the explanation Hadamard conversion.
Fig. 4 A, 4B and 4C are the schematic diagram of explanation intra prediction mode.
Fig. 5 is the flow chart in order to the processing of selecting an intra prediction mode of explanation according to first embodiment.
Fig. 6 is according to the use of first embodiment block diagram of the picture coding device of standard H.264.
Fig. 7 is the block diagram of the configuration of the image capture device of explanation a third embodiment in accordance with the invention.
Fig. 8 moves the schematic diagram of example of the intra prediction direction of the image that obtains in the process for the explanation camera lens.
Fig. 9 is explanation moves according to camera lens in the picture coding device of the 3rd embodiment or the direction of pitch rotation is selected intra prediction mode process chart.
Figure 10 is the schematic diagram that is illustrated in the example of the intra prediction direction that obtains image in the zoom process.
The process chart that Figure 11 selects intra prediction mode for the zoom in the picture coding device of explanation a fourth embodiment in accordance with the invention.
Figure 12 is the schematic diagram of the example of the infra-frame prediction of explanation blurred picture.
Figure 13 is the process chart of the intra prediction mode of the focusing in the explanation picture coding device according to a fifth embodiment of the invention.
Figure 14 is the block diagram according to the picture coding device of the 3rd embodiment.
Figure 15 is for illustrating the schematic diagram of the pixel of using in infra-frame prediction.
Figure 16 is the schematic diagram of nine kinds of intra prediction modes of explanation.
Figure 17 is the schematic diagram of the example of the direction of the infra-frame prediction of the common image of expression.
Figure 18 selects the flow chart of the common processing of intra prediction mode for explanation.
Embodiment
Embodiments of the present invention is described in detail with reference to the accompanying drawings.
First embodiment
Fig. 6 is the block diagram of expression according to the configuration of the picture coding device 600 of the first embodiment of the present invention.This picture coding device 600 comprises: subtracter 601, integer transform unit 602, quantization unit 603, entropy coding unit 604, inverse guantization (IQ) unit 605, inverse integer transform unit 606, adder 607, frame memory 608 and 611, intraprediction unit 609, loop filter 610, inter prediction unit 612, mobile detecting unit 613 and switch 614.Picture coding device 600 is configured to input image data is carried out the data of encoding process after with output encoder.
The following describes the encoding process in the picture coding device 600 described in Fig. 6.Picture coding device 600 bases H.264 standard are carried out encoding process.In addition, picture coding device 600 is divided into square with the moving image data of importing and each square is handled.
At first, subtracter 601 deducts the view data of prediction with the output differential image data from the view data (view data of input) that is input to picture coding device 600.The generation of predicted picture will be described in the back.
The 602 pairs of differential image data from subtracter 601 outputs in integer transform unit are carried out orthogonal transform with the output transform coefficient.Then, quantization unit 603 uses predetermined quantization parameter that conversion coefficient is carried out quantization.
Also be used for predicted image data that the square of back is encoded by quantization unit 603 quantized conversion coefficients as generating.Inverse guantization (IQ) unit 605 carries out the inverse guantization (IQ) operation on by quantization unit 603 quantized conversion coefficients.Inverse integer transform is carried out in inverse integer transform unit 606 on by the conversion coefficient after inverse guantization (IQ) unit 605 inverse guantization (IQ), with the difference data behind the output decoder.Adder 607 is with the view data addition of decoded difference data and prediction, with the view data after the output reconstruction.
Image data storage after the reconstruction stores in the frame memory 611 in frame memory 608 and by loop filter 610.The data predicted that may relate in the reconstructed image data afterwards temporarily is stored in frame memory 608 or 611.Loop filter 610 is used to remove barrage noise.
Fig. 1 is the block diagram of explanation according to the intra-prediction process of first embodiment.Configuration shown in Fig. 1 is corresponding to the intraprediction unit 609 that is included in the picture coding device 600.
With reference to figure 1, intraprediction unit 609 comprises picture pattern determining unit 101, intra prediction mode designating unit 102, selector 103, vertical intraprediction unit 104, horizontal intraprediction unit 105, direct current intraprediction unit 106 and selector 107.Thereby picture pattern determining unit 101 is determined picture pattern by carry out the Hadamard conversion on input picture.Intra prediction mode designating unit 102 is specified the intra prediction mode of an optimum from a plurality of intra prediction modes according to the picture pattern of being determined by picture pattern determining unit 101.Selector 103 is selected a unit with 107 from vertical intraprediction unit 104, horizontal intraprediction unit 105 and direct current intraprediction unit 106, corresponding to the intra prediction mode of intra prediction mode designating unit 102 appointments.Vertical intraprediction unit 104 uses the vertical frame inner estimation mode to carry out infra-frame prediction.Horizontal intraprediction unit 105 usage level intra prediction modes carry out infra-frame prediction.Direct current intraprediction unit 106 uses the direct current intra prediction mode to carry out infra-frame prediction.Like this, vertically, a predictive mode in level and the direct current intra prediction mode just is selected and is carried out according to the selection of selector 103 and 107.As a result of, intraprediction unit 609 prediction of output images.
Major part to the intraprediction unit shown in Fig. 1 609 is elaborated below.In the present embodiment, for example, picture pattern determining unit 101 is carried out the Hadamard conversion on the input picture that is divided into 4 * 4 block of pixels.Yet, be used for determining that the size of piece of picture pattern and configuration are not limited to these descriptions wherein.
Picture pattern determining unit 101 is divided into 4 * 4 block of pixels with the pixel data of input picture, on the pixel data of each piece, carry out the Hadamard conversion, and determine the picture pattern of each piece according to the Hadamard conversion coefficient that obtains by the Hadamard transformation calculations.
Below with reference to Fig. 2 the Hadamard conversion is described.The Hadamard conversion is a kind of orthogonal transform.The mode of 4 * 4 Hadamard conversion of Fig. 2 explanation in picture pattern determining unit 101.In Fig. 2, (a) part illustrates the input picture of being made up of 4 * 4 pixels, and (b) the part explanation is by the Hadamard conversion coefficient of Hadamard conversion acquisition.Suppose H
4Be the Hadamard transformation matrix, X is the picture signal of input, and Y is the signal of Hadamard conversion coefficient.So, the Hadamard conversion is represented with following equation:
[Y]=[H
4][X][H
4] (1)
Here, make
So, the Hadamard transform definition is:
Like this, the Hadamard conversion can only be finished with division, addition and a subtraction.Be noted that the left side Y of equation (5)
11The Hadamard conversion coefficient of the DC component of expression input picture, Y
12To Y
44The Hadamard conversion coefficient of the alternating current component of expression representing input images.
The example of the Hadamard conversion of Fig. 3 A, 3B and 3C explanation input picture, wherein the view data of each pixel 8bit (256 gray scale) is split into 4 * 4 block of pixels.In each Fig. 3 A, 3B and 3C, (a) part is represented input picture, and (b) part is represented the Hadamard conversion coefficient.Exist under the situation of vertical edges in the image block as the input that is shown in of Fig. 3 A (a) part, corresponding to Fig. 2 (b) the Hadamard conversion coefficient Y shown in partly
14The Hadamard conversion coefficient of position get a big value, shown in (b) part of Fig. 3 A.Exist under the situation of horizontal sides in the image block as the input that is shown in of Fig. 3 B (a) part, corresponding to Fig. 2 (b) the Hadamard conversion coefficient Y shown in partly
41The Hadamard conversion coefficient of position get a big value, shown in (b) part of Fig. 3 B.As shown in Fig. 3 C (a) part when the image block of input comprises plane picture, corresponding to the Hadamard conversion coefficient Y shown in (b) part of Fig. 2
11The Hadamard conversion coefficient of position get a big value, shown in (b) part of Fig. 3 C, other coefficient is got 0 value.Usually, exist under the situation of vertical edges in the image block of input, it is that low Hadamard conversion coefficient is got a big value that the spatial frequency on the indication vertical direction is only arranged, and encloses shown in the next Hadamard conversion coefficient with oval in (b) part of Fig. 3 A.Under horizontal sides was present in situation in the image block of input, it was that low Hadamard conversion coefficient is got a big value that the spatial frequency on the indication horizontal direction is only arranged, shown in the Hadamard conversion coefficient that encloses with ellipse in (b) part of Fig. 3 B.Comprise under the situation of plane picture in the image block of input, the Hadamard conversion coefficient of the alternating current component of indication input picture is got a little value.
The use said method makes such as the picture pattern of the input picture of vertical edges, horizontal sides or plane picture, can determine based on the Hadamard conversion coefficient that passes through execution Hadamard conversion acquisition on the image of input.
Below with reference to Fig. 4 A, 4B and 4C infra-frame prediction is described.Fig. 4 A, 4B and 4C illustrate the example of various intra prediction modes.
In Fig. 4 A, 4B and 4C, the pixel value of the input picture piece that a indicates to predict to p.A represents to belong to the pixel value of contiguous block to M.Pixel value a is arranged in same frame to p and A to M.Infra-frame prediction uses pixel value A to arrive p ' to the pixel value a ' of M generation forecast.The pixel value a ' of prediction concentrates in together the formation predicted picture to p '.
Below various intra prediction modes are described in detail.
In the vertical frame inner estimation mode shown in Fig. 4 A, predict in vertical direction.Equal pixel value A respectively by predicted pixel values a, e, i and m, pixel value b, f, j and n equal pixel value B respectively, pixel value c, g, k and o equal pixel value C respectively, and pixel value d, h, l and p equal pixel value D respectively, can arrive p ' by generation forecast pixel value a '.Like this, generate following pixel value:
a′＝e′＝i′＝m′＝A
b′＝f′＝j′＝n′＝B
c′＝g′＝k′＝o′＝C
d′＝h′＝l′＝p′＝D
In the horizontal intra prediction mode shown in Fig. 4 B, predict in the horizontal direction.Equal pixel value I respectively by predicted pixel values a, b, c and d, pixel value e, f, g and h equal pixel value J respectively, pixel value i, j, k and l equal pixel value K respectively, and pixel value m, n, o and p equal pixel value L respectively, thereby generation forecast pixel value a ' is to p '.Like this, generate following pixel value:
a′＝b′＝c′＝d′＝I
e′＝f′＝g′＝h′＝J
i′＝j′＝k′＝l′＝K
m′＝n′＝o′＝p′＝L
In the direct current intra prediction mode shown in Fig. 4 C, predict so that all pixels have same value.By predicting that all pixel value a equate that to p thereby generation forecast pixel value a ' is to p '.Like this, generate following pixel value:
a′＝b′＝c′＝d′＝e′＝f′＝g′＝h′＝i′＝j′＝k′＝l′＝m′＝n′＝o′＝p′
The predicted pixel values and the difference between the actual pixel value of input picture are more little, and it is high more that accuracy of predicting will become, and so just can carry out effective image compression.
Method according to picture pattern designated frame inner estimation mode is described below.
Intra prediction mode designating unit 102 is specified a kind of intra prediction mode of optimum according to the picture pattern of being determined by picture pattern determining unit 101 recited above from above-mentioned intra prediction mode.For example, if determine to comprise vertical edges in the input picture piece, then intra prediction mode designating unit 102 is specified the vertical frame inner estimation mode.If determine to comprise horizontal sides in the input picture piece, then intra prediction mode designating unit 102 specified level intra prediction modes.If determine to comprise plane picture in the input picture piece, intra prediction mode designating unit 102 is specified the direct current intra prediction mode.
More specifically, if vertical edges is present in the input picture piece as shown in Fig. 3 A the pixel value that then can obtain to have following relation:
a＝e＝i＝m
b＝f＝j＝n
c＝g＝k＝o
d＝h＝l＝p
This input picture piece is carried out the Hadamard conversion to obtain the Hadamard conversion coefficient.If picture pattern determining unit 101 determines that the input picture pieces have the picture pattern that comprises vertical edges, then intra prediction mode designating unit 102 is specified the vertical frame inner estimation modes, wherein accuracy of predicting for the input picture piece for the highest.
If as shown in Fig. 3 B, in the input picture piece, have horizontal sides, then obtain pixel value with following relation:
a＝b＝c＝d
e＝f＝g＝h
i＝j＝k＝l
m＝n＝o＝p
This input picture piece is carried out the Hadamard conversion to obtain the Hadamard conversion coefficient.If picture pattern determining unit 101 determines that the input picture pieces have the picture pattern that comprises horizontal sides, intra prediction mode designating unit 102 specified level intra prediction modes then, wherein accuracy of predicting for the input picture piece for the highest.
If when the input picture piece comprises plane picture as shown in Fig. 3 C, then obtain pixel value with following relation:
a＝b＝c＝d＝e＝f＝g＝h＝i＝j＝k＝l＝m＝n＝o＝p
This input picture piece is carried out the Hadamard conversion to obtain the Hadamard conversion coefficient.If picture pattern determining unit 101 determines that the input picture pieces have the picture pattern that comprises plane picture, then intra prediction mode designating unit 102 is specified the direct current intra prediction modes, wherein accuracy of predicting for the input picture piece for the highest.
Then, selector 103 and 107 makes the predicting unit of selecting carry out intra-prediction process on input picture from corresponding to selecting one the intraprediction unit 104 to 106 by the intra prediction mode of intra prediction mode designating unit 102 appointments like this.As a result of, can be according to the pixel value of optimal frames inner estimation mode generation forecast.In addition, if picture pattern determining unit 101 can not positively be determined the picture pattern of input picture and intra prediction mode designating unit 102 and can not specify the optimal frames inner estimation mode, if perhaps because the selected device 103 and 107 of neither one is selected in other reason intraprediction unit 104 to 107, then intraprediction unit 609 is exported this input picture as it is.
If consider in the process of determining picture pattern and carry out the required time of Hadamard conversion, the timing adjustment unit such as memory of the time-delay that then is used to provide predetermined, be introduced in selector 103 level before, perhaps intraprediction unit self can be carried out timing and adjusted, thereby for picture pattern and definite image can be complementary with the image that will carry out intra-prediction process in timing.
The processing sequence that is used to select an intra prediction mode below with reference to the flow chart description of the block diagram of Fig. 1 and Fig. 5 according to first embodiment.The flowchart text of Fig. 5 is corresponding to the handling process of control program, and the controller (not shown) is carried out this control program with each unit shown in the control chart 1.
At first, picture pattern determining unit 101 is divided into 4 * 4 block of pixels with input picture, and imports each block of pixels (step S501).Picture pattern determining unit 101 is carried out Hadamard conversion (step S502) then to determine picture pattern, and is as described below.
Picture pattern determining unit 101 determines whether 4 * 4 block of pixels of input comprise vertical edges (step S503), horizontal sides (step S504) or plane picture (step S505).
If picture pattern determining unit 101 determines that block of pixels comprises vertical edges (step S503 " yes "), then intra prediction mode designating unit 102 is specified the vertical frame inner estimation mode from a plurality of intra prediction modes, and selector 103 is selected vertical intraprediction unit 104 (step S506) with 107.
If picture pattern determining unit 101 determines that block of pixels comprises horizontal sides (" no " of step S503, and step S504 " yes "), then intra prediction mode designating unit 102 specified level intra prediction mode from a plurality of intra prediction modes, and selector 103 and 107 is selected horizontal intraprediction unit 105 (step S507).
If picture pattern determining unit 101 determines that block of pixels comprises plane picture (" no " of step S503, " no " of step S504, and step S505 " yes "), then intra prediction mode designating unit 102 is specified the direct current intra prediction mode from a plurality of intra prediction modes, and selector 103 and 107 is selected direct current intraprediction unit 106 (step S508).
If block of pixels does not have above-mentioned arbitrary picture pattern, that is to say, if picture pattern determining unit 101 be can not determine any picture pattern (being " no " among step S503, S504 and the S505), then selector 103 and 107 is not selected any one from intraprediction unit 104 to 107, and exports original input picture (step S509) same as before.Then, handling process finishes.
In addition, if picture pattern determining unit 101 be can not determine any picture pattern, then selector 103 and 107 can be configured to select intraprediction unit 104,105 or 106 independently.
In the first above-mentioned embodiment, vertical frame inner estimation mode, horizontal intra prediction mode and direct current intra prediction mode are used as intra prediction mode with selected.Yet, also can adopt above-mentioned intra prediction mode intra prediction mode in addition in the same way.
In addition, in the first above-mentioned embodiment, vertical edges, horizontal sides or plane picture are determined.Yet, thereby corresponding with it intra prediction mode is selected in the combination that also can determine these patterns.
In addition, in the first above-mentioned embodiment, one in vertical edges, horizontal sides and the plane picture is determined.Yet, also can do definitely like this, promptly reduce the intra prediction mode that will select.For example, determine that the input picture piece does not comprise any horizontal sides (that is to say that the input picture piece might comprise vertical edges or plane picture).Even in this configuration, can select optimum intra prediction mode effectively, and assessing the cost in picture coding device 600 can be reduced also.
In addition, in the first above-mentioned embodiment, picture pattern determining unit (Hadamard converter unit) 101 and intra prediction mode designating unit 102 are included in the intraprediction unit 609 of picture coding device 600.Yet picture pattern determining unit (Hadamard converter unit) 101 or intra prediction mode designating unit 102 can for example be positioned at beyond the picture coding device 600 or outside the intraprediction unit 609.For example, picture pattern determining unit 101 or intra prediction mode designating unit 102 can directly receive the input image data of the picture coding device 600 that will offer shown in Fig. 6, so that to determine picture pattern or designated frame inner estimation mode with the same mode of aforesaid way.
Second embodiment
The second embodiment of the present invention relates to the example in the modification of the picture pattern determining unit 101 described in first embodiment shown in Figure 1.In a second embodiment, picture pattern determining unit 101 is configured to use the method to input picture execution filtering to determine picture pattern, rather than uses the Hadamard transform method.The structure of the part except picture pattern determining unit 101 with the operation with first embodiment in describe identical, so omitted description of them here.
Picture pattern determining unit (filter unit) 101 is definite such as the picture pattern for a limit in a second embodiment, and this definite processing is to finish by the size of the intensity level on the limit of the input picture of each piece being used difference filter and estimation acquisition.Difference filter for using in picture pattern determining unit 101 for example, uses an operator.This operator is the matrix that wherein stores weighted value.Yet filtering is not limited to this configuration.
Below the frontier inspection survey method of using operator is described.
In using the frontier inspection survey method of operator, each weighted value is added on the pixel value of the object pixel of input picture and neighborhood pixels.Then, the pixel value after object pixel and the neighborhood pixels weighting is added together to constitute the limit intensity level of object pixel.The size that suitable weighted value is set and estimates the limit intensity level makes can determine picture pattern such as vertical edges, horizontal sides or plane picture.
When picture pattern is determined in the configuration of using picture pattern determining unit 101 as mentioned above and operation, the optimal frames inner estimation mode of picture pattern can with first embodiment in identical mode low assess the cost selected.
The 3rd embodiment
Figure 14 is the block diagram of the configuration of the picture coding device 1400 of expression a third embodiment in accordance with the invention.Picture coding device 1400 comprises subtracter 601, integer transform unit 602, quantization unit 603, entropy coding unit 604, inverse guantization (IQ) unit 605, inverse integer transform unit 606, adder 607, frame memory 608 and 611, loop filter 610, inter prediction unit 612, mobile detecting unit 613, switch 614 and intraprediction unit 615.Picture coding device 1400 is configured to input image data is carried out data behind encoding process and the output encoder.
The encoding process of the picture coding device 1400 shown in Figure 14 is described below.Picture coding device 1400 bases H.264 standard are carried out encoding process.In addition, picture coding device 1400 is divided into piece with input image data and each piece is handled.
In the block diagram of Figure 14, have and unit identical functions and operation described in Fig. 6 with the unit that has same reference numerals (601 to 608,610 to 614) in the block diagram of Fig. 6.In the following discussion, carry out repeat specification for fear of the unit described in Fig. 6, only to Fig. 6 in describe different points and be described.
In response to the control signal that the outside from picture coding device 1400 provides, in the 3rd embodiment, use the reconstructed image data that is stored in the frame memory 608 and carry out infra-frame prediction, with the generation forecast view data for unique intraprediction unit 615.Switch 614 is as selected cell, for each macro block is selected in infra-frame prediction and the inter prediction any.That is to say that switch 614 is selected one and the predicted image data of selecting outputed to subtracter 601 and adder 607 from the output of the output of intraprediction unit 615 and inter prediction unit 612.
Infra-frame prediction below with reference to Figure 15 and 16 pairs of intraprediction unit 615 execution replenishes description.
In infra-frame prediction, the value of the neighborhood pixels that before has been encoded is used as predicted value.If the object block that is used to encode as shown in Figure 15 is 4 * 4 pixel sizes (pixel a is to p), then use object block pixel A on every side to predict to the pixel value of M.In this case, the direction of prediction is designated and only have the pixel value that needs to be used as prediction.Because nine kinds of intra prediction modes that have different prediction direction and different content respectively are provided, so select the intra prediction mode of an optimum to be used as prediction.
Figure 16 represents to be used in the various prediction direction in each intra prediction mode.For example, under the situation of vertical frame inner estimation mode (pattern 0), the value of the pixel of adjacent objects piece (corresponding to the pixel A shown in Figure 15, B, C and D) is used as predicted value in vertical direction, and predicts that these pixel values continue in vertical direction.Therefore, the value of each pixel a, e, i and the m shown in prediction Figure 15 equals the value of pixel A.Equally, predict that the value of each pixel b, f, j and n equals the value of pixel B.Same prediction also is used for the pixel of back.After above-mentioned prediction, obtain different value, and then, the value that obtains is carried out integer transform and quantization etc. such as each pixel of a-A, e-A, i-A and m-A.
Except vertical frame inner estimation mode (pattern 0), provide eight kinds of intra prediction modes (pattern 1 to 8) as shown in Figure 16.Direct current intra prediction mode (pattern 2) is to use predictive mode higher and the mean value left pixel value.The details of each intra prediction mode is all disclosed in H.264 relevant standard etc., so, just omitted further description here.In standard H.264, always have nine kinds of intra prediction modes as mentioned above, so can select a kind of intra prediction mode of optimum for the target pixel block that will encode.Therefore, can improve accuracy of predicting, and can make the value of differential image data very little.
In addition, in intraprediction unit 615, select the common processing of optimal frames inner estimation mode to carry out with the order as shown in the flow chart of Figure 18.With reference to Figure 18, when beginning the infra-frame prediction of object block, provide the counter i and the counter i of expression intra prediction mode to be set to " 0 " (step S1801).Then, use is carried out infra-frame prediction (step S1802) by the intra prediction mode of the value representation of counter i.Precision of prediction in the estimated frames, and compare with intra prediction mode of having estimated before other and to select the intra prediction mode (step S1803) of indication precision of prediction optimal value.Then, counter i increases (step S1804).The order that repeats from step S1802 to step S1804 arrives " 9 " until counter i, and this numeral is the quantity (step S1805) of intra prediction mode.As a result of, each object block that will encode is carried out nine kinds of all intra prediction modes.Therefore, can select a kind of optimal frames inner estimation mode.
In addition, the control signal that response provides, intraprediction unit 615 also can be to handle the processing that different orders is selected the optimal frames inner estimation mode with above-mentioned common selection.This different selection is handled and be will be explained hereinafter.
Next in conjunction with the further operation of the intraprediction unit 615 relevant, the configuration of the image capture device 700 that comprises above-mentioned picture coding device 1400 is described with the operation of image capture device 700.
Fig. 7 is the block diagram of expression according to the configuration of the image capture device 700 of the 3rd embodiment.Image capture device 700 comprises phtographic lens unit 701, imageing sensor 702, photography video camera signal processing unit 703, coding unit 704, photography video camera control unit 705, focal length detecting unit 706, focusing motor 707, zoom motor 708, movable sensor 709 and operating unit 710.Phtographic lens unit 701 is made up of a plurality of lens group of zoom lens 701a and 701b and focusing lens 701c that comprise.Imageing sensor 702 is CCD (charge) transducer, CMOS (complementary metal oxide semiconductors (CMOS)) transducer etc.Operating unit 710 is exercisable for various functions.Image capture device 700 will be converted to picture signal by the optical imagery that phtographic lens unit 701 forms, and carry out compressed encoding according to standard H.264 on picture signal, thus outputting encoded data.In image capture device 700, coding unit 704 is corresponding to above-mentioned picture coding device 1400.Like this, image capture device 700 comprises the picture coding device 1400 as the H.264 encoder of video compression.
The photographic process of being undertaken by the image capture device shown in Fig. 7 700 is described below.When the beginning shooting operation, the optical imagery that is formed by phtographic lens unit 701 is converted to the signal of telecommunication by imageing sensor 702.In this case, focusing lens 701c can be moved focusing by focusing motor 707, and zoom lens 701a and 701b can be moved to carry out the adjustment of optical zoom by zoom motor 708.In addition, optical image stabilization mechanism (not shown) or electronic image stabilization function (not shown) can be operated so that to mobile correction the (for example, the vibration of photography video camera being proofreaied and correct) of image according to the mobile message that movable sensor 709 provides.
The signal of telecommunication that is generated by imageing sensor 702 is used as mobile picture signal and offers photography video camera signal processing unit 703.Photography video camera signal processing unit 703 carries out luminance signal processing, colour signal processing etc. to generate moving image data on mobile picture signal.Be provided to coding unit 704 from the moving image data of photography video camera signal processing unit 703 outputs.704 pairs of moving image datas of coding unit are encoded with outputting encoded data.In coding unit 704, carry out above-mentioned with reference to the described cataloged procedure of Figure 14.Data behind the coding are that a kind of unit (not shown) that can be recorded is recorded in such as the form on the recording medium of disk.
The operation of each unit of 705 pairs of image capture devices 700 of photography video camera control unit is controlled.If the user provides instruction by operating unit 710, then photography video camera control unit 705 is according to each unit of commands for controlling.
In addition, in the process of automatic focusing, focal length detecting unit 706 is according to the electrical signal detection focus state that is generated by imageing sensor 702, and the information (focusing information) of detected focus state is sent to photography video camera control unit 705.Photography video camera control unit 705 sends control signals to focusing motor 707 according to focusing information, so that carry out the focal length adjustment to obtain optimum focus state.In the process of manual focusing, photography video camera control unit 705 sends control signals to focusing motor 707 according to the instruction that provides from operating unit 710, so that carry out the focal length adjustment to obtain required focus state.
In addition, photography video camera control unit 705 sends control signals to zoom motor 708 according to the instruction that provides from operating unit 710, so that carry out in order to obtain the zoom control of required zoom position.
When movable sensor 709 detected pan or pitching, photography video camera control unit 705 carries out control operation, and was as described below.
The situation of the seizure image that obtains during the action of Fig. 8 explanation in the process of shooting operation, carrying out pan.If on the direction of the white arrow shown in Fig. 8 (not filling) picture is carried out pan, then the image of being caught shows as overall flow on the white direction of arrow (horizontal direction).So,, can estimate that then horizontal direction is the optimum prediction direction if infra-frame prediction is undertaken by coding unit 704 on the image of catching.In this case, can also estimate that as shown in Figure 8, the direction of the infra-frame prediction of each piece is aimed on the regular direction corresponding to panning direction, and under the situation of as shown in Figure 17 image, the direction of the infra-frame prediction of each piece is irregular on the picture.So, select the infra-frame prediction (horizontal intra prediction mode (pattern 1)) of horizontal direction to cause the optimum effectively selected or the intra prediction mode of near-optimization.If pitching in vertical direction then should be selected vertical frame inner estimation mode (pattern 0) in a similar fashion.In addition, if pan or pitching are carried out, should consider that then this slope component selects a diagonal angle intra prediction mode on tilted direction.
Like this, photography video camera control unit 705 sends control signals to the intraprediction unit 615 that is included in the coding unit 704 (that is, picture coding device 1400), in order to go out intra prediction mode according to the photography Action Selection such as pan or pitching.In addition, the control signal that response provides, intraprediction unit 615 are carried out the foregoing common processing in order to selection optimal frames inner estimation mode of following processing replacement.That is to say, intraprediction unit 615 is selected an intra prediction mode uniquely according to the direction of pan or pitching, perhaps selects a kind of optimal frames inner estimation mode according to the direction of pan or pitching after nine kinds of intra prediction modes are reduced to two or three intra prediction mode.Like this, can more effectively select the optimal frames inner estimation mode with the situation of selecting a kind of intra prediction mode than tentatively carrying out all intra prediction modes.
Fig. 9 is the flow chart of explanation according to the operating sequence of the intraprediction unit 615 of the direction selection intra prediction mode of above-mentioned pan or pitching.
If movable sensor 709 does not detect pan or pitching action (no of step S901), then intraprediction unit 615 is used to select the common processing of optimal frames inner estimation mode as shown in Figure 18, and just carries out infra-frame prediction (step S907) with the intra prediction mode of selecting.On the other hand, if movable sensor 709 detects about the pan of the image that will carry out infra-frame prediction or pitching (yes among the step S901), then intraprediction unit 615 obtains detected pan or the direction of pitching action or the information (step S902) of speed from photography video camera control unit 705.
According to the pan or the direction of pitching or the information of speed that obtain at step S902, intraprediction unit 615 is similar to selects prediction direction (step S903) automatically corresponding to the direction of pan or pitching.If S903 selects vertical direction in step, then intraprediction unit 615 uses vertical frame inner estimation mode (pattern 0) to carry out infra-frame prediction (step S904).If S903 selects horizontal direction in step, then intraprediction unit 615 usage level intra prediction modes (pattern 1) carry out infra-frame prediction (step S905).If select under the diagonal angle at step S903-right, then intraprediction unit 615 use down-right intra prediction mode (pattern 4) carries out infra-frame prediction (step S906).
As for to the prediction on the angular direction, can improve the directional precision of movable sensor 709 so that also can select other diagonal angle intra prediction mode (mode 3,5,6,7,8).Like this, accuracy of predicting just can further improve.
In addition, under the situation of the order shown in Fig. 9, when carrying out pan and pitching, intraprediction unit 615 is selected intra prediction mode according to the direction of pan or pitching.Yet, intraprediction unit 615 can be configured to by according to the direction of pan or pitching or speed nine kinds of intra prediction modes being reduced to two kinds or three kinds of intra prediction modes or are defined as obvious inappropriate intra prediction mode on prediction direction by eliminating, thereby reduce the option of intra prediction mode, and then, obtain the optimal frames inner estimation mode by from the option of remainder, calculating.
According to the 3rd above-mentioned embodiment, the information of moving such as the photography of pan or pitching is used for selecting intra prediction mode.Therefore, optimum intra prediction mode can be easily selected, and the processing burden on the coding can be reduced.
The 4th embodiment
Below with reference to Figure 10 and Figure 11 the fourth embodiment of the present invention is described.
In the 4th embodiment, when carrying out the zoom action at the image capture device described in the 3rd embodiment 700, photography video camera control unit 705 and intraprediction unit 615 are carried out following operation.
When the zoom that carries out zoom-up or zoom during shooting operation moves, on the directions of rays at picture center, showing in the mode that flows at the seizure image that obtains during the zoom.So, when coding unit 704 carries out infra-frame prediction on the image of catching, should consider the object block that will encode is positioned on which direction with respect to the center of picture.If carry out infra-frame prediction, then can improve the effect of prediction corresponding to this direction.
In this case, can estimate that zoom causes the direction of the infra-frame prediction shown in each piece of Figure 10, and under the situation of the image shown in Figure 17, the direction of the infra-frame prediction of each piece is irregular on the picture.More specifically, in the higher left side block on the picture in Figure 10, because image is to the center of upper reaches, angular direction to picture, so think down-right intra prediction mode (pattern 4) is optimum.Same, in the higher right side square on picture, think down-left intra prediction mode (mode 3) is optimum.Under the situation of the infra-frame prediction of standard H.264, the prediction samples that is used for infra-frame prediction is limited to the value of the contiguous square in the top or left side that is positioned at the target square.Therefore, in the low left side square on picture, think down-left intra prediction mode (mode 3) is optimum.Similarly, in the low right side square on picture, think down-right intra prediction mode (pattern 4) is optimum.
Like this, photography video camera control unit 705 sends control signals to the intraprediction unit 615 that is included in the coding unit 704 (that is, picture coding device 1400), so that go out an intra prediction mode according to the zoom Action Selection.In addition, in response to the control signal that provides, the processing below carrying out on the image that intraprediction unit 615 obtains during zoom replaces the common processing that the optimal frames inner estimation mode is selected in above-mentioned being used to.That is to say, intraprediction unit 615 is selected a kind of intra prediction mode uniquely according to the position of the target square that will encode, perhaps selects the optimal frames inner estimation mode according to the position of described target square after nine kinds of intra prediction modes are reduced to two or three intra prediction mode.Like this, can more effectively select the optimal frames inner estimation mode with the situation of selecting optimal prediction modes than tentatively carrying out all intra prediction modes.
Figure 11 is used to select flow chart about the operating sequence of the intraprediction unit 615 of the intra prediction mode of the image that obtains as described above for explanation during zoom.
As for the image that will carry out infra-frame prediction, if do not carry out zoom action (if or carried out the zoom action of special low speed) (no among the step S1101), then intraprediction unit 615 is carried out the common processing that is used to select the optimal frames inner estimation mode as shown in Figure 18, and uses the intra prediction mode of selecting to carry out infra-frame prediction (step S1106).On the other hand, if response is carried out the zoom action with zoom lens motionless zoom lens 701a and 701b (yes among the step S1101) from the zoom instructions of operating unit 710, then intraprediction unit 615 obtains the information (step S1102) of indication ongoing information of zoom and zoom rate from photography video camera control unit 705.
Obtain the information of indication ongoing information of zoom and zoom rate at step S1102 after, intraprediction unit 615 is determined the zone at the object block place that will encode on picture, and selects a prediction direction (step S1103) automatically corresponding to the zone of determining.More specifically, if object block is positioned at higher left field or low right side area on the picture, then intraprediction unit 615 is selected down-right intra prediction mode (pattern 4).If object block is positioned at higher right side area or low left field, then intraprediction unit 615 is selected down-left intra prediction mode (mode 3).If object block is positioned at around central area or its, then intraprediction unit 615 can not the designated frame inner estimation mode.So intraprediction unit 615 carries out carrying out infra-frame prediction (step S1106) in order to the intra prediction mode that the common processing of selecting the optimal frames inner estimation mode and use are selected.If under in step S1103, selecting-and left intra prediction mode (mode 3), then intraprediction unit 615 uses following-left intra prediction mode (mode 3) to carry out infra-frame prediction (step S1104).If under step S1103 selects-and right intra prediction mode (pattern 4), then intraprediction unit 615 uses following-right intra prediction mode (pattern 4) to carry out infra-frame prediction (step S1105).
In addition, if picture is divided into a plurality of zones meticulously and is classified in the zone that obtains, then can more effectively carry out the selection of optimal frames inner estimation mode.For example, if object block be positioned at the picture center directly over or under, then intraprediction unit 615 is selected vertical frame inner estimation modes (pattern 0).If object block is positioned at the left side or the right side at picture center, then intraprediction unit 615 is selected horizontal intra prediction mode (pattern 1).In addition, make that other diagonal angle intra prediction mode (pattern 5,6,7,8) is also selected, then can further improve accuracy of predicting if picture is divided into a plurality of zones thinlyyer.
In addition, under the situation of the order shown in Figure 11, when carrying out zoom, intraprediction unit 615 is according to the choice of location intra prediction mode of the object block that will encode, unless around object block is positioned at central area or its.Yet, thereby intraprediction unit 615 also can be configured to by nine kinds of intra prediction modes being reduced to two kinds or three kinds of intra prediction modes or being defined as the selection that on prediction direction obvious inappropriate intra prediction mode reduces intra prediction mode by eliminating, and afterwards, from the selection of remainder, obtain the optimal frames inner estimation mode by calculating.
According to the 4th above-mentioned embodiment, be used for selecting intra prediction mode such as the photography action message of zoom.Therefore, can easily select optimal frames inner estimation mode and also can reduce processing burden on the coding.
The 5th embodiment
Be described below with reference to Figure 12 and 13 pairs of fifth embodiment of the present invention.
In the 5th embodiment, when the image capture device of describing among the 3rd embodiment 700 is focused when action, photography video camera control unit 705 and intraprediction unit 615 be as described below carries out work.
When focusing when action during shooting operation, and when not obtaining in focus state, the whole image of the image of seizure is blured.Therefore, when coding unit 704 carries out infra-frame prediction on fuzzy seizure image, can not estimate the concrete direction of prediction, and the direct current intra prediction mode of use pixel value average is considered to improve the optimal frames inner estimation mode of forecasting efficiency.
In this case, can estimate, the fringe of catching image causes the direct current prediction case of whole image as shown in Figure 12, and under the situation at state in focus shown in Figure 17, the direction of the infra-frame prediction of each square on the picture is irregular.
Like this, photography video camera control unit 705 sends control signals to the intraprediction unit 615 that is included in the coding unit 704 (that is, picture coding device 1400) in order to select intra prediction mode according to focus state.In addition, the control signal that provides of response, intraprediction unit 615 is selected the direct current intra prediction mode uniquely for fuzzy seizure image, is substituted by to select the optimal frames inner estimation mode and carry out above-mentioned common processing.Like this, thus can select the situation of intra prediction mode more effectively select the optimal frames inner estimation mode than tentatively carrying out all intra prediction mode.
Figure 13 is that explanation is used to select the flow chart about the operating sequence of the intraprediction unit 615 of the intra prediction mode of above-mentioned fuzzy image, the image about catching especially under the state of focus alignment.
Intraprediction unit 615 from photography video camera control unit 705 obtain indicating the image that will carry out infra-frame prediction be state in focus or under the focus alignment state captive focusing information (step S1301).At step S1302, if determine that from described focusing information image is being captured under the state in focus, then intraprediction unit 615 is carried out the common processing shown in Figure 18 for selecting the optimal frames inner estimation mode, and uses the intra prediction mode of selecting to carry out infra-frame prediction (step S1304).
At step S1302, if determining image from described focusing information is captured under the focus alignment state, be that image blurs, then intraprediction unit 615 is automatically selected the direct current intra prediction mode and is used direct current intra prediction mode (pattern 2) to carry out infra-frame prediction (step S1303).
According to the 5th above-mentioned embodiment, be used for selecting intra prediction mode such as the photography action message of focusing.Therefore, can easily select optimal frames inner estimation mode and can reduce processing burden on the coding.
Operating sequence described in the 3rd embodiment to the five embodiment can carry out maybe can combining carrying out independently.Carry out if these operations combine, then can respond various photography conditions and select the optimal frames inner estimation mode effectively and also can reduce processing burden on the coding.
Each embodiment also can realize by the system or the device that have storage medium are provided, store the functional programs code (software) that is used for realizing the foregoing description in this storage medium, the computer (or CPU (CPU), MPU (microprocessing unit) etc.) that this program code impels this system or device is from the storage medium program code read and carry out this program code then.In this case, the program code of reading from storage medium itself is realized the function of described embodiment.In addition, except carrying out the program code that reads by computer, present invention resides in according to the instruction of program code and intactly or partly carry out actual treatment so that the operating system (OS) that realizes moving on the computer of function of the foregoing description etc. to realize the function of the foregoing description.The example of program code stored storage medium comprises floppy disk, hard disk, ROM (read-only memory), RAM (random access memory), tape, Nonvolatile memory card, CD-ROM (compact disk-read-only memory), CD-R (compression recordable disc), DVD (digital versatile dish), Blu-ray Disc, CD, MO (magneto optical disk) etc.
In addition, the present invention also comprises being included in and is inserted into the function expansion card in the computer or is connected to CPU in the functional expansion unit on the computer etc., described function expansion card or functional expansion unit have thereon the internal memory with the program code that reads from storage medium, and described CPU etc. intactly or partly carry out actual treatment according to the instruction of program code so that realize the function of the foregoing description.
If the present invention is applied in the above-mentioned storage medium, then the program code corresponding to above-mentioned flow chart is stored in this storage medium.
Though the present invention is described with reference to the embodiment of example, be to be understood that the present invention is not limited to disclosed embodiment.On the contrary, the invention is intended to cover various modifications and equivalent in the spirit and scope that are included in appended claim.The scope of following claim is consistent with the wideest explanation so that comprise all this modifications and equivalent configurations and function.
Claims (19)
1. picture coding device comprises:
Input unit is configured to import the view data that is divided into image block, and each image block comprises a plurality of pixels;
Intraprediction unit is configured to according to a plurality of intra prediction modes predicted pixel values in described image block;
Converter unit is configured to carry out orthogonal transform about pixel value in being included in each image block of described input image data; And
Selected cell is configured to select the optimal frames inner estimation mode among a plurality of intra prediction modes of each image block,
Wherein said selected cell is selected an optimal frames inner estimation mode based on the result of the described orthogonal transform of described converter unit, and wherein said intraprediction unit is according to described intra prediction mode predicted pixel values in described image block of being selected by described selected cell, and the view data of the prediction of output.
2. according to the picture coding device of claim 1, wherein said converter unit is configured to determine in vertical edges, horizontal sides and the plane at least one by carrying out described orthogonal transform for each image block.
3. according to the picture coding device of claim 1 or 2, wherein said selected cell is configured to select different intra prediction modes according to the result of the described orthogonal transform of described converter unit.
4. according to the picture coding device of claim 1, wherein said converter unit is configured to carry out the Hadamard conversion as described orthogonal transform.
5. according to the picture coding device of claim 4, wherein said selected cell is configured to select one from following processing, promptly, when being big numerical value, the Hadamard conversion coefficient of the low vertical spatial frequency of indication selects the processing of vertical prediction pattern, when the Hadamard conversion coefficient of the low-level spatial frequency of indication is selected the processing of horizontal forecast pattern when being big numerical value, and when the Hadamard conversion coefficient of indication alternating current component during for fractional value selection make all pixels have the processing of the predictive mode of identical value.
6. method for encoding images may further comprise the steps:
Input is divided into the view data of image block, and each image block comprises a plurality of pixels;
According to a plurality of intra prediction modes predicted pixel values in described image block;
In being included in each image block of described input image data, carry out orthogonal transform about pixel value; And
Based on the result of described orthogonal transform, among a plurality of intra prediction modes of each image block, select the optimal frames inner estimation mode,
Wherein said prediction steps is according to described intra prediction mode predicted pixel values in described image block of selecting in described selection step, and the view data of the prediction of output.
7. picture coding device comprises:
Input unit is configured to import the view data that is divided into image block, and each image block comprises a plurality of pixels;
Intraprediction unit is configured to according to a plurality of intra prediction modes predicted pixel values in described image block;
Acquiring unit is configured to obtain to indicate the photographic information of the shooting operation of image capture device when described view data is captured; And
Selected cell is configured to one of a plurality of intra prediction modes of selecting each image block,
Wherein said selected cell is selected corresponding to by the described intra prediction mode by the indicated described shooting operation of the described photographic information that described acquiring unit obtained, and wherein said intraprediction unit is according to described intra prediction mode predicted pixel values in described image block of being selected by described selected cell, and the view data of the prediction of output.
8. according to the picture coding device of claim 7, wherein said selected cell is selected among n the intra prediction mode corresponding to by one of m candidate's predictive mode of the indicated described shooting operation of described photographic information, wherein 1≤m≤n.
9. according to the picture coding device of claim 7 or 8, wherein said photographic information comprises the information that is obtained by acceleration sensor, the information of zoom action and a kind of information in the information of moving of focusing at least.
10. according to the picture coding device of claim 7 or 8, wherein said photographic information comprises the information that the described view data of indication is captured under the condition in pan, pitching, zoom and image fog condition at least.
11. picture coding device according to claim 7 or 8, wherein said acquiring unit is configured to obtain the photographic information that moves of indication selected shooting direction when described view data is captured, and described selected cell is configured to use the mobile message of described shooting direction to select the intra prediction mode of indication corresponding to the prediction direction of described moving direction.
12. picture coding device according to claim 7 or 8, wherein said acquiring unit is configured to obtain to indicate the photographic information of zoom performed when having caught described view data, and described selected cell is configured to use zoom information to select predetermined intra prediction mode when the image of being caught during described image data table is shown in zoom.
13. picture coding device according to claim 12, wherein, when the image of being caught during the described pictorial data representation zoom, it is a plurality of zones that described selected cell is configured to picture segmentation, and selects the described intra prediction mode that has adopted same prediction direction in each zone.
14. picture coding device according to claim 7 or 8, wherein said acquiring unit is configured to obtain to indicate the photographic information of focusing performed when having caught described view data, and described selected cell is configured to use focusing information to select predetermined intra prediction mode when described image data table is shown in the image of catching under the focus alignment state.
15. the picture coding device according to claim 7 also comprises:
Camera unit is configured to and will offers described input unit by the view data of carrying out the shooting operation acquisition; And
The photographic information generation unit is configured to generate photographic information when described camera unit is carried out shooting operation, and photographic information is offered described acquiring unit.
16. picture coding device according to claim 15, wherein said photographic information generation unit is configured to use from the information that detects the transducer output that described camera unit moves and generates photographic information, and, according to this photographic information, described selected cell is configured to select the intra prediction mode of indication corresponding to the prediction direction of the moving direction of described camera unit when described pictorial data representation is used the image of the camera unit seizure of moving.
17. picture coding device according to claim 15, wherein, described photographic information generation unit is configured to generate photographic information according to the information of control zoom action in described camera unit, and, according to this photographic information, described selected cell is configured to select predetermined intra prediction mode when described image data table is shown in the image of being caught during zoom moves.
18. picture coding device according to claim 15, wherein, described photographic information generation unit is configured to generate photographic information according to the information of control focusing action in described camera unit, and, according to this photographic information, described selected cell is configured to select predetermined intra prediction mode when described image data table is shown in the image of catching under the focus alignment state.
19. a method for encoding images may further comprise the steps:
Input is divided into the view data of image block, and each image block comprises a plurality of pixels;
According to a plurality of intra prediction modes predicted pixel values in described image block;
Obtain the photographic information of indication shooting operation of image capture device when described view data is captured; And
Select one of a plurality of intra prediction modes of each image block,
Wherein said selection step is selected corresponding to the described intra prediction mode by the indicated described shooting operation of described photographic information, and wherein said infra-frame prediction step is according to described intra prediction mode predicted pixel values in described image block of selecting in described selection step, and the view data of the prediction of output.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2004179933A JP2006005659A (en) | 2004-06-17 | 2004-06-17 | Device and method for image encoding |
| JP179933/2004 | 2004-06-17 | ||
| JP026530/2005 | 2005-02-02 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1710962A CN1710962A (en) | 2005-12-21 |
| CN100393137C true CN100393137C (en) | 2008-06-04 |
Family
ID=35707124
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100772084A Expired - Fee Related CN100393137C (en) | 2004-06-17 | 2005-06-16 | Moving image coding apparatus |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP2006005659A (en) |
| CN (1) | CN100393137C (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4609670B2 (en) * | 2004-12-28 | 2011-01-12 | 日本電気株式会社 | Image encoding apparatus, image encoding method and program thereof |
| FR2908007A1 (en) * | 2006-10-31 | 2008-05-02 | Thomson Licensing Sas | Image sequence i.e. video, encoding method for video compression field, involves pre-selecting intra-encoding modes of macro-block with directions that correspond to gradient activities with minimal value to determine encoding mode |
| US8331448B2 (en) | 2006-12-22 | 2012-12-11 | Qualcomm Incorporated | Systems and methods for efficient spatial intra predictabilty determination (or assessment) |
| US8428133B2 (en) | 2007-06-15 | 2013-04-23 | Qualcomm Incorporated | Adaptive coding of video block prediction mode |
| CN101690235B (en) * | 2007-06-29 | 2013-05-01 | 夏普株式会社 | Image encoding device, image encoding method, image decoding device, image decoding method, program, and recording medium |
| JP5085808B2 (en) * | 2011-01-21 | 2012-11-28 | パナソニック株式会社 | Moving picture coding apparatus and moving picture coding method |
| PL3739884T3 (en) * | 2011-06-27 | 2021-11-22 | Samsung Electronics Co., Ltd. | Encoding and decoding motion information |
| JP2013179572A (en) * | 2012-02-08 | 2013-09-09 | Canon Inc | Image encoding apparatus and image encoding method |
| CN103297776B (en) * | 2013-05-17 | 2016-02-10 | 西安电子科技大学 | Non-registration multispectral image lossless/lossy coding System and method on star |
| CN109215062B (en) * | 2017-06-29 | 2022-02-08 | 沈阳新松机器人自动化股份有限公司 | Motion capture method based on image vision, binocular positioning device and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1238888A (en) * | 1996-11-27 | 1999-12-15 | 普林斯顿视频图像公司 | Image insertion in video streams using combination of physical sensors and pattern recognition |
| JP2001145011A (en) * | 1999-11-15 | 2001-05-25 | Victor Co Of Japan Ltd | Video signal encoder |
| CN1535027A (en) * | 2004-01-16 | 2004-10-06 | 北京工业大学 | Inframe prediction method used for video frequency coding |
-
2004
- 2004-06-17 JP JP2004179933A patent/JP2006005659A/en active Pending
-
2005
- 2005-06-16 CN CNB2005100772084A patent/CN100393137C/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1238888A (en) * | 1996-11-27 | 1999-12-15 | 普林斯顿视频图像公司 | Image insertion in video streams using combination of physical sensors and pattern recognition |
| JP2001145011A (en) * | 1999-11-15 | 2001-05-25 | Victor Co Of Japan Ltd | Video signal encoder |
| CN1535027A (en) * | 2004-01-16 | 2004-10-06 | 北京工业大学 | Inframe prediction method used for video frequency coding |
Non-Patent Citations (8)
| Title |
|---|
| Fast Mode Decision for Intra Prediction. Feng PAN ET AL.JVT of ISO/IEC MPEG & ITU-T VCEG 7th Meeting:Pattaya,Thailand,7-14 March,2003,JVT-G013. 2003 |
| Fast Mode Decision for Intra Prediction. Feng PAN ET AL.JVT of ISO/IEC MPEG & ITU-T VCEG 7th Meeting:Pattaya,Thailand,7-14 March,2003,JVT-G013. 2003 * |
| Fast Mode Decision for Intra Prediction. Feng PAN ET AL.JVT of ISO/IEC MPEG & * |
| Hardware Architecture Design for H.264/AVC IntraFrameCoder. Yu-Wen Huang ET AL.Proceedings of the 2004 International Symposium on Circuits and Systems, 2004. ISCAS '04,Vol.2 . 2004 |
| Hardware Architecture Design for H.264/AVC IntraFrameCoder. Yu-Wen Huang ET AL.Proceedings of the 2004 International Symposium on Circuits and Systems, 2004. ISCAS '04,Vol.2 . 2004 * |
| ITU-T VCEG 7th Meeting:Pattaya,Thailand,7-14 March,2003,JVT-G013. 2003 * |
| Multistage Mode Decision for Intra Prediction in H.264Codec. Changsung Kim et al.Proceedings of Visual Communications and Image Processing 2004,SPIE,Bellingham,VA,US,Vol.5308 No.1. 2004 |
| Multistage Mode Decision for Intra Prediction in H.264Codec. Changsung Kim et al.Proceedings of Visual Communications and Image Processing 2004,SPIE,Bellingham,VA,US,Vol.5308 No.1. 2004 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1710962A (en) | 2005-12-21 |
| JP2006005659A (en) | 2006-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100393137C (en) | Moving image coding apparatus | |
| EP1770639B1 (en) | Moving image coding apparatus | |
| CN103650504B (en) | Control based on image capture parameters to Video coding | |
| JP5261376B2 (en) | Image coding apparatus and image decoding apparatus | |
| JP4591657B2 (en) | Moving picture encoding apparatus, control method therefor, and program | |
| CN104320664B (en) | Image processing equipment and method | |
| KR101339566B1 (en) | Motion vector selection | |
| JP6149707B2 (en) | Moving picture coding apparatus, moving picture coding method, moving picture coding program, and moving picture photographing apparatus | |
| US11206418B2 (en) | Method of image encoding and facility for the implementation of the method | |
| JP2008227670A (en) | Image coding device | |
| TW202015403A (en) | Video coding and decoding | |
| JP2010057166A (en) | Image coding apparatus, image coding method, integrated circuit, and camera | |
| JP4898415B2 (en) | Moving picture coding apparatus and moving picture coding method | |
| JP4885698B2 (en) | Motion vector calculation device, method and program, moving image compression recording device, and imaging device | |
| JP4564856B2 (en) | Image encoding apparatus and imaging apparatus | |
| JPH10327401A (en) | Motion vector detection method and encoding method and device for picture signal using the same | |
| JP2006295734A (en) | Re-encoding apparatus, re-encoding method, and re-encoding program | |
| JP7507912B2 (en) | Method for encoding and decoding an image, encoding and decoding device and corresponding computer program product | |
| JP7507913B2 (en) | Method for encoding and decoding an image, encoding and decoding device and corresponding computer program product | |
| JP7148236B2 (en) | Encoding device and its control method and program | |
| JP2009049513A (en) | Moving picture encoding device and moving picture encoding method | |
| JP2009021786A (en) | Image encoding device and method | |
| US20130201367A1 (en) | Image encoding apparatus and its control method | |
| JP6016484B2 (en) | Encoder | |
| CN112740673A (en) | Image data encoding and decoding |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080604 Termination date: 20210616 |