Image prediction method and system
A method is provided for computing a predicted frame from a first and a second reference frames, the method comprising, for each block of pixels in the predicted frame, the acts of defining a first block of pixels in the first reference frame collocated with a third block of pixels which is the block of pixels in the predicted frame; defining a second block of pixels corresponding, in the second reference frame, to the first block of pixels along the motion vector of the first block from the first to second reference frames; computing a first set of coefficients allowing the transformation of the pixels of the first block into pixels of the second block; computing pixels of the third block using the first set of coefficients and pixels from a fourth block collocated in the first reference frame with the second block of pixels.
1. A method for computing a predicted frame Y t from a first reference frame X t-1 and a second reference frame X t-2 , said method comprising, for each block of pixels in the predicted frame Y t :
a) defining a first block B t-1 (k,l) of pixels in the first reference frame X t-1 , wherein a third block B t (k,l) of pixels in the predicted frame Y t is in the same location in the predicted frame Y t as the location of the first block B t-1 (k,l) in the first reference frame X t-1 ,
b) defining a second block B t-2 ({tilde over (k)}, {tilde over (l)}) of pixels corresponding, in the second reference frame X t-2 , to the first block B t-1 (k,l) of pixels along a motion vector v t-1,t-2 (k,l) of said first block B t-1 (k,l) from said first reference frame X t-1 to said second reference frame X t-2 ,
c1) computing, using a processor, a first set of interpolation coefficients α i,j allowing a transformation of the pixels of the first block B t-1 (k,l) into pixels of the second block B t-2 ({tilde over (k)}, {tilde over (l)}),
c2) computing, using the processor, a second set of interpolation coefficients β i,j allowing the transformation of the pixels of the second block B t-2 ({tilde over (k)}, {tilde over (l)}) into pixels of the first block B t-1 (k,l),
d) computing, using a processor, a first prediction of the pixels of the third block B t (k,l) using the first set of interpolation coefficients α i,j and a second prediction of the pixels from a fourth block B t-1 ({tilde over (k)}, {tilde over (l)}) collocated in the first reference frame X t-1 with the second block B t-2 ({tilde over (k)}, {tilde over (l)}) of pixels; and
e) deriving, using the processor, the optimum prediction of the pixels in of third block B t (k,l) as being an average of the first and second predictions.
2. The method according to claim 1 , wherein the blocks of pixels are square blocks of n×n pixels, with n being an integer greater than 1, the first and second set of interpolation coefficients corresponding to a first and second n×n matrices, and wherein in d), the computing takes into account said first matrix and the transposition of said second matrix.
3. An interpolating device for computing a predicted frame from a first and a second reference frames of a video flow, said device being arranged to select said first and second frames from the video flow, said device being further arranged for each block of pixels in the predicted frame to:
a) define a first block B t-1 (k,l) of pixels in the first reference frame X t-1 , wherein a third block B t (k,l) of pixels in the predicted frame Y t is in the same location in the predicted frame Y t as the location of the first block B t-1 (k,l) in the first reference frame X t-1 ,
b) define a second block B t-2 ({tilde over (k)}, {tilde over (l)}) of pixels corresponding, in the second reference frame X t-2 , to the first block B t-1 (k,l) of pixels along a motion vector v t-1,t-2 (k,l) of said first block B t-1 (k,l) from said first reference frame X t-1 to said second reference frame X t-2 ,
c1) compute a first set of interpolation coefficients α i,j allowing a transformation of the pixels of the first block B t-1 (k,l) into pixels of the second block B t-2 ({tilde over (k)}, {tilde over (l)}),
c2) compute a second set of interpolation coefficients β i,j allowing the transformation of the pixels of the second block B t-2 ({tilde over (k)}, {tilde over (l)}) into pixels of the first block B t-1 (k,l),
d1) compute a first prediction of the pixels of the third block B t (k,l) using the first set of interpolation coefficients α i,j and a second prediction of the pixels from a fourth block B t-1 ({tilde over (k)}, {tilde over (l)}) collocated in the first reference frame X t-1 with the second block B t-2 ({tilde over (k)}, {tilde over (l)}) of pixels; and
e) derive the optimum prediction of the pixels in of third block B t (k,l) as being an average of the first and second predictions.
4. The device according to claim 3 , wherein the blocks of pixels are square blocks of n×n pixels, with n being an integer greater than 1, the first and second set of interpolation coefficients corresponding to a first and second n×n matrices, the device being further arranged to compute the pixels of the predicted frame block taking into account said first matrix and the transposition of said second matrix.
5. A system for computing a predicted frame from a first and a second reference frame of a video flow, said system comprising:
a transmitting device for transmitting the video flow,
an interpolating device arranged to:
receive the video flow from the transmitting device, and
select said first and second frames from the video flow,
said device being further arranged for, for each block of pixels in the predicted frame, to:
a) define a first block B t-1 (k,l) of pixels in the first reference frame X t-1 , wherein a third block B t (k,l) of pixels in the predicted frame Y t is in the same location in the predicted frame Y t as the location of the first block B t-1 (k,l) in the first reference frame X t-1 ,
b) define a second block B t-2 ({tilde over (k)}, {tilde over (l)}) of pixels corresponding, in the second reference frame X t-2 , to the first block B t-1 (k,l) of pixels along a motion vector v t-1,t-2 (k,l) of said first block B t-1 (k,l) from said first reference frame X t-1 to said second reference frame X t-2 ,
c1) compute a first set of interpolation coefficients α i,j allowing a transformation of the pixels of the first block B t-1 (k,l) into pixels of the second block B t-2 ({tilde over (k)}, {tilde over (l)}),
c2) compute a second set of interpolation coefficients β i,j allowing the transformation of the pixels of the second block B t-2 ({tilde over (k)}, {tilde over (l)}) into pixels of the first block B t-1 (k,l),
d1) compute a first prediction of the pixels of the third block B t (k,l) using the first set of interpolation coefficients α i,j and a second prediction of the pixels from a fourth block B t-1 ({tilde over (k)}, {tilde over (l)}) collocated in the first reference frame X t-1 with the second block B t-2 ({tilde over (k)}, {tilde over (l)}) of pixels; and
e) derive the optimum prediction of the pixels in of third block B t (k,l) as being an average of the first and second predictions.
6. The system according to claim 5 , wherein the blocks of pixels are square blocks of n×n pixels, with n being an integer greater than 1, the first and second set of interpolation coefficients corresponding to a first and second n×n matrices, the device being further arranged to compute the pixels of the predicted frame block taking into account said first matrix and the transposition of said second matrix.
7. A non-transitory computer program product providing computer executable instructions stored on a computer readable medium, which when loaded on to a data processor causes the data processor to perform the method for computing a predicted frame from a first and a second reference frames according to claim 1 .