CN101964881A - Method and system for realizing image rotation during digital television reception - Google Patents
Method and system for realizing image rotation during digital television reception Download PDFInfo
- Publication number
- CN101964881A CN101964881A CN 200910158216 CN200910158216A CN101964881A CN 101964881 A CN101964881 A CN 101964881A CN 200910158216 CN200910158216 CN 200910158216 CN 200910158216 A CN200910158216 A CN 200910158216A CN 101964881 A CN101964881 A CN 101964881A
- Authority
- CN
- China
- Prior art keywords
- frame
- module
- residual error
- original picture
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/60—Rotation of a whole image or part thereof
- G06T3/606—Rotation by memory addressing or mapping
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/387—Composing, repositioning or otherwise geometrically modifying originals
- H04N1/3877—Image rotation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/44—Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/61—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/85—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
Abstract
The invention provides a method and a system for realizing image rotation during digital television reception. The method comprises the following steps that: an entropy decoding module receives and decodes a video elementary stream to acquire a discrete cosine transform (DCT) transform domain matrix of a block or a block residue and a motion vector; an inverse quantization module performs weighting inverse quantization on the DCT transform domain matrix and performs inverse transformation on each coefficient of the DCT transform domain matrix to a value before quantization; an inverse DCT module converts the DCT transform domain matrix after the weighting inverse quantization into a space domain matrix; a matrix transposition module performs matrix transposition operation on an original image block or an original image block residue and stores the transposed image block serving as one part of a reconstruction frame into a frame memory; and the frame memory stores a display frame, the reconstruction frame and a predictive frame. By the method, rotation display can be performed on the image after video decoding under the condition of not increasing access times and storage space.
Description
Technical field
The present invention relates in Digital Television receives,, realize the method and system of image rotation as mobile TV, terrestrial broadcasting etc.
Background technology
Digital Television is considered to one of the most wide industry of 21 century prospect, it also is the emphasis of digital consumer electronic information products development in the international coverage, Digital Television receives and has been penetrated into various terminals at present, comprises mobile phone, PMP, PSP, PC, set-top box or the like.China domestic terrestrial broadcasting standard and mobile TV standard are also established.
Block-based dct transform (discrete cosine transform) algorithm is widely used in comprising MPEG1/2/4 in the various video standards, H.26x with AVS in.Fig. 1 is a typical digital TV decoding schematic diagram based on the piece conversion.
When decoded picture showed, owing to watch the difference of custom, the user generally needed image is rotated demonstration, especially to mobile phone or PMP screen.
The implementation method that traditional rotation shows need be from the display frame buffer memory reading of data, rotation is deposited at last in the display frame buffer memory and (is likely another section zone), as shown in Figure 2 then.Because decoded image data amount is bigger, such realization will increase the access times to memory, even needs to increase the requirement to storage size, thereby causes the increase of terminal price, power consumption to increase.
Summary of the invention
The technical problem to be solved in the present invention provides a kind of method and system of realizing the image rotation in Digital Television receives, realizes the image behind the video decode is rotated demonstration under the situation that does not increase access times and memory space.
According to an aspect of the present invention, propose a kind of system that in Digital Television receives, realizes the image rotation, comprising: the entropy decoder module, reception and decoded video flow substantially, obtain the discrete cosine transform dct transform domain matrix and the motion vector of piece or piece residual error; Inverse quantization module, the dct transform domain matrix is weighted inverse quantization, numerical value before quantizing is changed in each coefficient contravariant of dct transform domain matrix, through the coefficient of the dct transform domain matrix behind the weighting inverse quantization corresponding to coding side image block or piece residual error through the result behind the dct transform; Anti-DCT module is the spatial domain matrix with the dct transform domain matrix conversion behind the weighting inverse quantization, promptly is converted to original picture block or the original picture block residual error of 4x4 or 8x8; The matrix transpose module is carried out the matrix transpose computing to original picture block or original picture block residual error, and image block behind the transposition is deposited into frame memory as the part of reconstruction frames; Frame memory, storage display frame, reconstruction frames and predictive frame, reconstruction frames is transformed to predictive frame in decode procedure, and when the number of predictive frame in the decode procedure reached set point, predictive frame will become display frame.
Further, the system of described realization image rotation also comprises: the motion vector correcting module, according to the anglec of rotation of setting, the original motion vector in the entropy decoder module revised obtain the vector that rotatablely moves, and the vector that will rotatablely move sends in the frame/the inter prediction module; In the frame/the inter prediction module, when type of prediction is infra-frame prediction, the pixel in a zone that reads the vectors directed that rotatablely moves from frame memory obtains reconstructed block with the original picture block residual error addition behind predict pixel and the transposition, and reconstructed block is write frame memory as predict pixel.
Further, in the frame/the inter prediction module is also when type of prediction is inter prediction, read from the predictive frame position of frame memory the vectors directed that rotatablely moves a zone pixel as the prediction piece, original picture block residual error addition behind prediction piece and the transposition is obtained reconstructed block, and reconstructed block is write frame memory.
According to a further aspect of the invention, a kind of method that realizes the image rotation in Digital Television receives is also proposed, may further comprise the steps: entropy decoder module receiver, video flows substantially, video-frequency basic flow is decoded, obtain the discrete cosine transform dct transform domain matrix and the motion vector of piece or piece residual error; Inverse quantization module will be weighted inverse quantization to the dct transform domain matrix, numerical value before quantizing is changed in each coefficient contravariant of dct transform domain matrix, through the coefficient of the dct transform domain matrix behind the weighting inverse quantization corresponding to coding side image block or piece residual error through the result behind the dct transform; The dct transform domain matrix conversion of anti-DCT module after with the weighting inverse quantization is the spatial domain matrix, promptly is converted to original picture block or the original picture block residual error of 4x4 or 8x8; The matrix transpose module is carried out the matrix transpose computing to original picture block or original picture block residual error, and original picture block behind the transposition is deposited into frame memory as the part of reconstruction frames; Frame memory storage display frame, reconstruction frames and predictive frame, reconstruction frames are transformed to predictive frame in decode procedure, when the number of predictive frame in the decode procedure reached set point, predictive frame will become display frame.
Further, after the original picture block residual error carried out the matrix transpose computing, further comprising the steps of: according to the anglec of rotation of setting, the motion vector correcting module is revised the original motion vector in the entropy decoder module and is obtained the vector that rotatablely moves, and the vector that will rotatablely move sends in the frame/the inter prediction module; When type of prediction is infra-frame prediction, in the frame/the inter prediction module reads the pixel in a zone of the vectors directed that rotatablely moves as predict pixel from frame memory; Original picture block residual error addition behind predict pixel and the transposition is obtained reconstructed block, reconstructed block is write in the reconstruction frames of frame memory.
Further, after the original picture block residual error carried out the matrix transpose computing, further comprising the steps of: according to the anglec of rotation of setting, the motion vector correcting module is revised the original motion vector in the entropy decoder module and is obtained the vector that rotatablely moves, and the vector that will rotatablely move sends in the frame/the inter prediction module; When type of prediction is inter prediction, in the frame/the inter prediction module reads the pixel in a zone of the vectors directed that rotatablely moves as the prediction piece from the predictive frame position of frame memory; Original picture block residual error addition behind prediction piece and the transposition is obtained reconstructed block, reconstructed block is write in the reconstruction frames of frame memory.
Compared with prior art, the present invention has eliminated the extra demand of memory being read number of times, has eliminated the extra demand to storage space, and has reduced power consumption.
Description of drawings
The system schematic of video decode during the Digital Television that Figure 1 shows that prior art receives.
Figure 2 shows that the system schematic of the realization image rotation of prior art.
Figure 3 shows that the present invention realizes the system schematic of image block rotation.
The present invention shown in Figure 4 realizes the system schematic of image block residual error rotation.
Figure 5 shows that 8 * 8 matrixes are through the embodiment of 90 degree transposition among the present invention.
Fig. 6 (a) is depicted as the preceding relative position schematic diagram of rotation.
Fig. 6 (b) is depicted as rotation back relative position schematic diagram.
Figure 7 shows that the present invention realizes the method flow diagram to the image block rotation.
Figure 8 shows that the present invention realizes the method flow diagram that the image block residual error is rotated by infra-frame prediction.
The image block that Figure 9 shows that a 4x4 when doing infra-frame prediction and the schematic diagram of the position of neighbor relation, a wherein, b, c, d, e, f, g, h, I, j, k, l, m, n, o, p are each pixel of described 4x4 image block, ABCD is 4 pixels of the below of image block top adjacent image piece for this reason, and IJKL is 4 the most right-hand pixels of image block left adjacent image piece for this reason, and M is the pixel of the last cell of image block upper left side adjacent image piece for this reason.
Figure 10 shows that the present invention realizes the method flow diagram that the image block residual error is rotated by inter prediction.
Embodiment
The system schematic of video decode during the Digital Television that Figure 1 shows that prior art receives comprises in entropy decoder module 10, anti-DCT module 20, inverse quantization module 30, the frame/inter prediction module 40 and frame memory 50.
The video-frequency basic flow of entropy decoder module 10 other modules of receiving digital television receiving system obtains video-frequency basic flow after with digital TV transmission stream TS demultiplexing and audio frequency flows substantially as demultiplexing module, and wherein video-frequency basic flow is as the input of entropy decoding 10.Digital TV receiving system directly obtains video-frequency basic flow from hard disk or network for another example, as the input of entropy decoder module 10.Described digital TV receiving system can be a digital TV set-top box, comprises ground, satellite, wired or network.Also can be the portable terminal that has the digital TV decoding function, as mobile phone, PMP etc.Described video-frequency basic flow ES is based on the compressed bit stream of DCT piece conversion, as H.264, MPEG1/2/4, AVS etc.The decoding algorithm of entropy decoder module 10 foundations can be huffman coding, as MPEG1/2/3, or CABIC, CAVLC, as H.264 waiting.The dct transform domain matrix that obtains can be that 4x4 (as H.264) also can be 8x8 (as MPEG2).
30 pairs of described dct transform domain matrixes of inverse quantization module are weighted inverse quantization, each coefficient contravariant of dct transform domain matrix are changed to quantize preceding numerical value.Through the coefficient of the dct transform domain matrix behind the inverse quantization corresponding to coding side image block or piece residual error through the result behind the dct transform.
The dct transform domain matrix conversion of anti-DCT module 20 after with the weighting inverse quantization is the spatial domain matrix, promptly is converted to original picture block or the original picture block residual error of 4x4 or 8x8.
When being converted to original picture block, described original picture block is sent to frame memory 50, to make up reconstruction frames 502.When being converted to the original picture block residual error, then obtain predicting piece according to type of prediction, described prediction piece and the addition of original picture block residual error are obtained reconstructed block.Wherein, type of prediction comprises in the frame, inter prediction, the operation of obtaining the prediction piece by in the frame/inter prediction module 40 carries out, will be described below.
In the frame/40 pairs of original picture block residual errors of inter prediction module predict and obtain predicting piece, will predict that piece and the addition of original picture block residual error obtain reconstruction frames 502.
When type of prediction is infra-frame prediction, in the frame/and reconstruction frames 502 that inter prediction module 40 is finished from decoding obtains corresponding predict pixel, and this predict pixel will obtain according to the relevant position (as the boundary pixel of adjacent image macro) of infra-frame prediction type from reconstruction frames 502.The respective pixel addition obtains reconstructed block in the image block residual error that this predict pixel and anti-DCT module 20 obtain, and this reconstructed block is write the relevant position of reconstruction frames 502 in the frame memory 50.With standard H.264 is example, if the infra-frame prediction type is a 4x4 luminance frame inner estimation mode 1, it is vertical prediction, then in the frame/obtain 4 pixels of the below of described image block top image block the reconstruction frames 502 that inter prediction module 40 will be finished from decoding, ABCD as shown in Figure 9, and with these four pixels and current image block residual error (a, b, c, d; E, f, g, h; I, j, k, l; M, n, o, p) vertical summation obtains reconstructed block for (a+A, b+B, c+C, d+D; E+A, f+B, g+C, h+D; I+A, j+B, k+C, l+D; M+A, n+B, o+C p+D), writes this reconstructed block the relevant position of reconstruction frames 502 in the frame memory 50.
When type of prediction is inter prediction, in the frame/inter prediction module 40 is according to the motion vector that obtains from entropy decoder module 10, predictive frame 503 positions are predicted piece accordingly from frame memory 50, and this prediction piece will obtain pixel according to the relevant position of inter prediction type from predictive frame 503 and also calculate through filtering.Suppose that the inter prediction type is the P frame, it is forward prediction, and motion vector be (A, B), and described image block residual error is (k in the position of reconstruction frames 502 (first location of pixels of piece), l), then in the frame/inter prediction module 40 will from predictive frame 503 (k+A, an image block is read in position B+l), promptly predicts piece, the image block residual error addition that this prediction piece and anti-DCT module 20 obtain obtains reconstructed block, this reconstructed block is write the relevant position of reconstruction frames 502 in the frame memory 50.
Figure 2 shows that the system schematic of the realization image rotation of prior art.On the basis of Fig. 1, also comprise twiddle operation module 60, display frame 501 is carried out rotary manipulation.
After in Fig. 1 process, obtaining being arranged in the display frame 501 of frame memory 50, twiddle operation module 60 is read display frame 501 from frame memory 50, according to the anglec of rotation of setting whole display frame 501 is done twiddle operation, operation result write back to becomes new display frame 501 in the frame memory 50, and this display frame is carried out reading displayed.
Further describe operation principle of the present invention below in conjunction with drawings and Examples.
Figure 3 shows that the present invention realizes compared to Figure 1 the system schematic of image block rotation has been increased matrix transpose module 80.
Frame memory, storage display frame, reconstruction frames and predictive frame, reconstruction frames is transformed to predictive frame in decode procedure, and when the number of predictive frame in the decode procedure reached set point, predictive frame will become display frame.
It is before making up display frame that image block is carried out the twiddle operation operation, has avoided the operation of reading displayed frame from frame memory, has eliminated frame memory is read number of times and to the extra demand of memory space.
Because this moment, this system was used for image block is carried out rotary manipulation, therefore, can include only entropy decoder module 10, anti-DCT module 20, inverse quantization module 30 and frame memory 50 in the prior art system.Certainly, this operation also can be based on also comprise in the frame/inter prediction module 40 system finished.
Figure 4 shows that the present invention realizes the system schematic of image block residual error rotation, also comprises motion vector correcting module 70 on the basis of Fig. 3.Other modules below in conjunction with prior art specify the operation that this module is carried out.
Motion vector correcting module 70 is according to the anglec of rotation of setting, the original motion vector in the entropy decoder module 10 is revised is obtained the vector that rotatablely moves, and will the described vector that rotatablely moves send to frame interior/inter prediction module 40.
In the frame/when inter prediction module 40 is infra-frame prediction in type of prediction, the pixel in a zone that reads the vectors directed that rotatablely moves from frame memory reconstruction frames 502 is as predict pixel, original picture block residual error addition behind this predict pixel and the transposition obtains reconstructed block, this reconstructed block is write the relevant position of reconstruction frames 502 in the frame memory 50.It is consistent that prior art reads mode with content of the present invention in this part, and difference is prior art the saying of vector that do not rotatablely move, the original motion vector that be based on obtains from code stream.
With standard H.264 is example, if the infra-frame prediction type is a 4x4 luminance frame inner estimation mode 1, it is vertical prediction, then in the frame/obtain 4 pixels of the below of described image block top image block the reconstruction frames 502 that inter prediction module 40 will be finished from decoding, ABCD as shown in Figure 9, and with these four pixels and current image block residual error (a, b, c, d; E, f, g, h; I, j, k, l; M, n, o, p) vertical summation obtains reconstructed block for (a+A, b+B, c+C, d+D; E+A, f+B, g+C, h+D; I+A, j+B, k+C, l+D; M+A, n+B, o+C p+D), writes this reconstructed block the relevant position of reconstruction frames 502 in the frame memory 50.
In the frame/when inter prediction module 40 is inter prediction in type of prediction, from frame memory 50 predictive frame 503 positions read the vectors directed that rotatablely moves a zone pixel as the prediction piece, original picture block residual error addition behind prediction piece and the transposition is obtained reconstructed block, and reconstructed block is write frame memory.This prediction piece will obtain pixel according to the relevant position of inter prediction type from predictive frame 503 and also calculate through filtering.
Fig. 6 (a) is depicted as the embodiment schematic diagram of the preceding relative position of rotation, supposes that the inter prediction type is P frame, i.e. forward prediction, decoded motion vector is (A, B), and described image block residual error in the position of reconstruction frames 502 (first location of pixels of piece) be (k, l), then the position of former reference image block is (X, Y), X=k+A wherein, Y=b+l, and the size of supposing image for (H, V).Fig. 6 (b) is depicted as and revolves the embodiment schematic diagram that turn 90 degrees the back relative position.When revolve turn 90 degrees after, motion vector will be corrected for (B, A), revised reference block locations is (V-Y, X), promptly (V-B-l, k+A), in the frame/inter prediction module 40 will be from (the V-b-l of predictive frame 503, k+A) image block is read in the position, promptly predict piece, the image block residual error addition behind this prediction matrix and the transposition obtains reconstructed block, this reconstructed block is write the relevant position of reconstruction frames 502 in the frame memory 50.
The correction of rotating other angles in like manner can get, and concrete modification method is as follows:
The Rotate 180 degree
Motion vector be corrected for (A ,-B), revised reference block locations be (H-X, V-Y), promptly (H-k-A, V-B-l).
Rotation 270 degree
Motion vector be corrected for (B ,-A), revised reference block locations be (Y, H-X), promptly (B+l, H-A-k).
Owing to the predict pixel that obtains according to reconstruction frames 502 and according to the prediction piece that predictive frame 503 obtains all is the matrix transpose computing of having passed through matrix transpose module 80, and 70 outputs of motion vector correcting module also is to revise the vector that rotatablely moves that obtains.Simultaneously, be the image block residual error of rotation through what obtain behind matrix transpose module 80 transposition, so, the reconstructed block that predict pixel or prediction piece and the addition of image block residual error obtain, comparing with original image is postrotational image.
The present invention is under the situation that does not increase frame memory 50 capacity and reading times, in the decode procedure of prior art, increase matrix transpose module 80, realization is to the rotary manipulation and the demonstration of image block, further, motion vector correcting module 70 be can also increase, rotary manipulation and demonstration realized the image block residual error.
Figure 7 shows that the present invention realizes the method flow diagram of image block rotation be may further comprise the steps:
In step 701, entropy decoder module 10 receiver, videos flow substantially, and described video-frequency basic flow is decoded, and obtain the discrete cosine transform dct transform domain matrix and the motion vector of piece or grand residual error.
In step 702,30 pairs of described dct transform domain matrixes of inverse quantization module are weighted inverse quantization, numerical value before quantizing is changed in each coefficient contravariant of dct transform domain matrix, through the coefficient of the dct transform domain matrix behind the weighting inverse quantization corresponding to coding side image block or piece residual error through the result behind the dct transform.
In step 703, the dct transform domain matrix conversion of anti-DCT module 20 after with the weighting inverse quantization is the spatial domain matrix, promptly is converted to original picture block or the original picture block residual error of 4x4 or 8x8.
In step 704, the original picture block or the described original picture block residual error of 80 pairs of anti-DCT module 20 outputs of matrix transpose module are carried out the matrix transpose computing, the part of image block behind the transposition as reconstruction frames is deposited in the reconstruction frames 502 of frame memory 50, reconstruction frames 502 is transformed to predictive frame 503 in decode procedure, use for the decoding forecasting institute in future.When the number of predictive frame in the decode procedure reached set point, predictive frame will become display frame.
Described matrix transpose computing can be rotation 0,90,180,270 degree or mirror image etc.When the number of predictive frame in the decode procedure 503 reached set point, promptly this decode procedure no longer needed a certain predictive frame 503, and this predictive frame 503 will become display frame 501, and is used for showing.Shown image promptly is to the postrotational image of original image.
Because it is before making up display frame that image block is carried out the twiddle operation operation, has avoided the operation of reading displayed frame from frame memory, has eliminated frame memory has been read number of times and to the extra demand of memory space.
Figure 8 shows that the present invention realizes method flow diagram to the rotation of image block residual error by infra-frame prediction, on the basis of step 701～step 704 of Fig. 7, further comprising the steps of:
In step 801, according to the anglec of rotation of setting, the original motion vector in 70 pairs of entropy decoder modules 10 of motion vector correcting module is revised and is obtained the vector that rotatablely moves, and will the described vector that rotatablely moves send to frame interior/inter prediction module 40.In step 802, when type of prediction is infra-frame prediction, in the frame/inter prediction module 40 reads the pixel in a zone of the vectors directed that rotatablely moves as predict pixel from the reconstruction frames 502 of frame memory 50.
In step 803, the image block residual error addition behind predict pixel and the transposition is obtained reconstructed block, this reconstructed block is write the relevant position of reconstruction frames 502 in the frame memory 50.It is consistent that prior art reads mode with content of the present invention in this part, and difference is prior art the saying of vector that do not rotatablely move, the original motion vector that be based on obtains from code stream.
Figure 10 shows that the present invention realizes method flow diagram to the rotation of image block residual error by inter prediction, on the basis of step 701～step 704 of Fig. 7, further comprising the steps of:
In step 1001, according to the anglec of rotation of setting, the original motion vector in 70 pairs of entropy decoder modules 10 of motion vector correcting module is revised and is obtained the vector that rotatablely moves, and will the described vector that rotatablely moves send to frame interior/inter prediction module 40.
In step 1002, when type of prediction is inter prediction, in the frame/inter prediction module 40 reads the pixel in a zone of the vectors directed that rotatablely moves as the prediction piece from predictive frame 503 positions of frame memory 50.
In step 1003, the original picture block residual error addition behind prediction piece and the transposition is obtained reconstructed block, reconstructed block is write in the reconstruction frames 502 of frame memory 50.
Owing to the predict pixel that obtains according to reconstruction frames 502 and according to the prediction piece that predictive frame 503 obtains all is the matrix transpose computing of having passed through matrix transpose module 80, and 70 outputs of motion vector correcting module also is to revise the vector that rotatablely moves that obtains.Simultaneously, be the image block residual error of rotation through what obtain behind matrix transpose module 80 transposition, so, the reconstructed block that predict pixel or prediction piece and the addition of image block residual error obtain, comparing with original image is postrotational image.
Claims (6)
1. in Digital Television receives, realize the system that image rotates for one kind, comprising:
The entropy decoder module receives and decoded video flows substantially, obtains the discrete cosine transform dct transform domain matrix and the motion vector of piece or piece residual error;
Inverse quantization module, the dct transform domain matrix is weighted inverse quantization, numerical value before quantizing is changed in each coefficient contravariant of dct transform domain matrix, through the coefficient of the dct transform domain matrix behind the weighting inverse quantization corresponding to coding side image block or piece residual error through the result behind the dct transform;
Anti-DCT module is the spatial domain matrix with the dct transform domain matrix conversion behind the weighting inverse quantization, promptly is converted to original picture block or the original picture block residual error of 4x4 or 8x8;
The matrix transpose module is carried out the matrix transpose computing to original picture block or original picture block residual error, and image block behind the transposition is deposited into frame memory as the part of reconstruction frames;
Frame memory, storage display frame, reconstruction frames and predictive frame, reconstruction frames is transformed to predictive frame in decode procedure, and when the number of predictive frame in the decode procedure reached set point, predictive frame will become display frame.
2. according to the system of the described realization image rotation of claim 1, also comprise:
The motion vector correcting module according to the anglec of rotation of setting, is revised the original motion vector in the entropy decoder module and to be obtained the vector that rotatablely moves, and the vector that will rotatablely move sends in the frame/the inter prediction module;
In the frame/the inter prediction module, when type of prediction is infra-frame prediction, the pixel in a zone that reads the vectors directed that rotatablely moves from frame memory obtains reconstructed block with the original picture block residual error addition behind predict pixel and the transposition, and reconstructed block is write frame memory as predict pixel.
3. according to the system of the described realization image of claim 2 rotation, wherein, in the frame/the inter prediction module is also when type of prediction is inter prediction, read from the predictive frame position of frame memory the vectors directed that rotatablely moves a zone pixel as the prediction piece, original picture block residual error addition behind prediction piece and the transposition is obtained reconstructed block, and reconstructed block is write frame memory.
4. in Digital Television receives, realize the method that image rotates for one kind, may further comprise the steps:
Entropy decoder module receiver, video flows substantially, and video-frequency basic flow is decoded, and obtains the discrete cosine transform dct transform domain matrix and the motion vector of piece or piece residual error;
Inverse quantization module will be weighted inverse quantization to the dct transform domain matrix, numerical value before quantizing is changed in each coefficient contravariant of dct transform domain matrix, through the coefficient of the dct transform domain matrix behind the weighting inverse quantization corresponding to coding side image block or piece residual error through the result behind the dct transform;
The dct transform domain matrix conversion of anti-DCT module after with the weighting inverse quantization is the spatial domain matrix, promptly is converted to original picture block or the original picture block residual error of 4x4 or 8x8;
The matrix transpose module is carried out the matrix transpose computing to original picture block or original picture block residual error, and original picture block behind the transposition is deposited into frame memory as the part of reconstruction frames;
Frame memory storage display frame, reconstruction frames and predictive frame, reconstruction frames are transformed to predictive frame in decode procedure, when the number of predictive frame in the decode procedure reached set point, predictive frame will become display frame.
5. according to the method for the described realization image of claim 4 rotation, the original picture block residual error carried out the matrix transpose computing after, further comprising the steps of:
According to the anglec of rotation of setting, the motion vector correcting module is revised the original motion vector in the entropy decoder module and is obtained the vector that rotatablely moves, and the vector that will rotatablely move sends in the frame/the inter prediction module;
When type of prediction is infra-frame prediction, in the frame/the inter prediction module reads the pixel in a zone of the vectors directed that rotatablely moves as predict pixel from frame memory;
Original picture block residual error addition behind predict pixel and the transposition is obtained reconstructed block, reconstructed block is write in the reconstruction frames of frame memory.
6. according to the method for the described realization image of claim 4 rotation, the original picture block residual error carried out the matrix transpose computing after, further comprising the steps of:
According to the anglec of rotation of setting, the motion vector correcting module is revised the original motion vector in the entropy decoder module and is obtained the vector that rotatablely moves, and the vector that will rotatablely move sends in the frame/the inter prediction module;
When type of prediction is inter prediction, in the frame/the inter prediction module reads the pixel in a zone of the vectors directed that rotatablely moves as the prediction piece from the predictive frame position of frame memory;
Original picture block residual error addition behind prediction piece and the transposition is obtained reconstructed block, reconstructed block is write in the reconstruction frames of frame memory.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200910158216 CN101964881A (en) | 2009-07-21 | 2009-07-21 | Method and system for realizing image rotation during digital television reception |
| PCT/CN2010/001079 WO2011009289A1 (en) | 2009-07-21 | 2010-07-16 | Method and system for implementing image rotation in reception of digital television |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200910158216 CN101964881A (en) | 2009-07-21 | 2009-07-21 | Method and system for realizing image rotation during digital television reception |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101964881A true CN101964881A (en) | 2011-02-02 |
Family
ID=43498731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200910158216 Pending CN101964881A (en) | 2009-07-21 | 2009-07-21 | Method and system for realizing image rotation during digital television reception |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101964881A (en) |
| WO (1) | WO2011009289A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013063784A1 (en) * | 2011-11-03 | 2013-05-10 | Thomson Licensing | Video encoding and decoding based on image refinement |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09322165A (en) * | 1996-05-31 | 1997-12-12 | Sony Corp | Image decoding device and method, and image reproducing device |
| JP2004336451A (en) * | 2003-05-08 | 2004-11-25 | Matsushita Electric Ind Co Ltd | Image decoding unit, image encoder and encoding method employing it, and image decoder and decoding method |
| US7643694B2 (en) * | 2004-12-31 | 2010-01-05 | Zoran Corporation | Method and apparatus for processing a compressed image in an order other than the order in which it was compressed |
| TW200726260A (en) * | 2005-12-22 | 2007-07-01 | Sunplus Technology Co Ltd | Block decoding method and device capable of longitudinally decoding and outputting |
| CN100578605C (en) * | 2007-04-12 | 2010-01-06 | 深圳安凯微电子技术有限公司 | An image rotation processing method, device and multi-media processor |
-
2009
- 2009-07-21 CN CN 200910158216 patent/CN101964881A/en active Pending
-
2010
- 2010-07-16 WO PCT/CN2010/001079 patent/WO2011009289A1/en active Application Filing
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013063784A1 (en) * | 2011-11-03 | 2013-05-10 | Thomson Licensing | Video encoding and decoding based on image refinement |
| US9930329B2 (en) | 2011-11-03 | 2018-03-27 | Thomson Licensing | Video encoding and decoding based on image refinement |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2011009289A1 (en) | 2011-01-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102792689B (en) | Delta compression can be carried out and for by image, remote display is presented to the amendment of estimation and metadata | |
| JP6120390B2 (en) | Predicted characteristics compensated for next-generation video content | |
| CN103283238B (en) | Utilization carries out the method and apparatus encoding and decoding according to the parallel infra-frame prediction of coding unit | |
| CN102783149B (en) | For the adaptive motion resolution of video coding | |
| CN101248673B (en) | Deblock filtering techniques for video coding according to multiple video standards | |
| CN101394569B (en) | Error tolerance method, device and chip for AVS video decoder | |
| CN102342108B (en) | Image Processing Device And Method | |
| CN104125468A (en) | Image processing device and method | |
| CN104380740A (en) | Encoding device, encoding method, decoding device, and decoding method | |
| KR20170056849A (en) | Method of encoding video, video encoder performing the same and electronic system including the same | |
| US20120207211A1 (en) | Image encoding and decoding apparatus and method | |
| CN103004196A (en) | Inclusion of switched interpolation filter coefficients in a compressed bit-stream | |
| CN105491389A (en) | 4X4 transform for media coding | |
| CN1615025A (en) | Apparatus capable of performing both block-matching motion compensation and global motion compensation and method thereof | |
| CN102160381A (en) | Image processing device and method | |
| CN103997650A (en) | Video decoding method and video decoder | |
| JP5260757B2 (en) | Moving picture encoding method, moving picture decoding method, moving picture encoding apparatus, and moving picture decoding apparatus | |
| US20070133689A1 (en) | Low-cost motion estimation apparatus and method thereof | |
| EP4218240A1 (en) | Template matching prediction for versatile video coding | |
| CN103535041A (en) | Image processing device and method | |
| CN102396232A (en) | Image-processing device and method | |
| WO2018089146A1 (en) | Conversion buffer to decouple normative and implementation data path interleaving of video coefficients | |
| US8879629B2 (en) | Method and system for intra-mode selection without using reconstructed data | |
| TW202234883A (en) | Intra block copy scratch frame buffer | |
| TWI455587B (en) | Circuit and method for multi-format video codec |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20110202 |