System and method for two-pass interpolation for quarter-pel motion compensation
View Patent ↗Methods and systems for processing a video signal are disclosed herein. Pixels may be interpolated utilizing at most two passes. During a first of the two passes, an aligned sub-pixel value may be interpolated utilizing a plurality of integer pixel values, if the aligned sub-pixel value is horizontally or vertically aligned with the plurality of integer pixel values. During a second of the two passes, a non-aligned sub-pixel value may be interpolated utilizing the interpolated aligned sub-pixel value, if the non-aligned sub-pixel value is not horizontally or vertically aligned with the plurality of integer pixel values. The aligned sub-pixel value may be interpolated utilizing 4-tap filtering of the integer pixel values.
1. A method for processing a video signal, the method comprising: using a processor to perform the following steps:
interpolating pixels utilizing at most two passes, wherein,
during a first of the two passes,
interpolating at least one aligned sub-pixel value utilizing a plurality of integer pixel values, if the at least one aligned sub-pixel value is horizontally or vertically aligned with the plurality of integer pixel values; and
during a second of the two passes,
interpolating at least one non-aligned sub-pixel value utilizing the interpolated at least one aligned sub-pixel value, if the at least one non-aligned sub-pixel value is not horizontally or vertically aligned with the plurality of integer pixel values.
2. The method according to claim 1 , comprising interpolating the at least one aligned sub-pixel value utilizing 4-tap filtering of the integer pixel values.
3. The method according to claim 1 , comprising interpolating the at least one aligned sub-pixel value utilizing 4-tap filtering of the integer pixel values, wherein the 4-tap filtering of the integer pixel values comprises performing at least one of the following filtering operations:
f 1( x, y, z, w )=(−5 x +56 y +15 z −2 w +32)>>6;
f 2( x, y, z, w )=(−5 x +37 y +37 z −5 w +32)>>6; and
f 3( x, y, z, w )=(−2 x +15 y +56 z −5 w +32)>>6.
4. The method according to claim 1 , comprising interpolating the at least one non-aligned sub-pixel value utilizing 4-tap filtering of the interpolated at least one aligned sub-pixel value.
5. The method according to claim 1 , comprising interpolating the at least one non-aligned sub-pixel value utilizing 4-tap filtering of the interpolated at least one aligned sub-pixel value, wherein the 4-tap filtering of the interpolated at least one aligned sub-pixel value comprises performing at least one of the following filtering operations:
f 1( x, y, z, w )=(−5 x +56 y +15 z −2 w +32)>>6;
f 2( x, y, z, w )=(−5 x +37 y +37 z −5 w +32)>>6; and
f 3( x, y, z, w )=(−2 x +15 y +56 z −5 w +32)>>6.
6. The method according to claim 1 , comprising interpolating the at least one aligned sub-pixel value utilizing at least one of a plurality of polyphase filters to filter the integer pixel values.
7. The method according to claim 6 , comprising interpolating the at least one non-aligned sub-pixel value utilizing at least one of a plurality of polyphase filters to filter the interpolated at least one aligned sub-pixel value.
8. A computer-readable storage medium having stored thereon, a computer program having at least one code section for processing a video signal, the at least one code section being executable by a computer for causing the computer to perform steps comprising:
interpolating pixels utilizing at most two passes, wherein,
during a first of the two passes,
interpolating at least one aligned sub-pixel value utilizing a plurality of integer pixel values, if the at least one aligned sub-pixel value is horizontally or vertically aligned with the plurality of integer pixel values; and
during a second of the two passes,
interpolating at least one non-aligned sub-pixel value utilizing the interpolated at least one aligned sub-pixel value, if the at least one non-aligned sub-pixel value is not horizontally or vertically aligned with the plurality of integer pixel values.
9. The computer-readable storage medium according to claim 8 , comprising code for interpolating the at least one aligned sub-pixel value utilizing 4-tap filtering of the integer pixel values.
10. The computer-readable storage medium according to claim 8 , comprising code for interpolating the at least one aligned sub-pixel value utilizing 4-tap filtering of the integer pixel values, wherein the 4-tap filtering of the integer pixel values comprises performing at least one of the following filtering operations:
f 1( x, y, z, w )=(−5 x +56 y +15 z −2 w +32)>>6;
f 2( x, y, z, w )=(−5 x +37 y +37 z −5 w +32)>>6; and
f 3( x, y, z, w )=(−2 x +15 y +56 z −5 w +32)>>6.
11. The computer-readable storage medium according to claim 8 , comprising code for interpolating the at least one non-aligned sub-pixel value utilizing 4-tap filtering of the interpolated at least one aligned sub-pixel value.
12. The computer-readable storage medium according to claim 8 , comprising code for interpolating the at least one non-aligned sub-pixel value utilizing 4-tap filtering of the interpolated at least one aligned sub-pixel value, wherein the 4-tap filtering of the interpolated at least one aligned sub-pixel value comprises performing at least one of the following filtering operations:
f 1( x, y, z, w )=(−5 x +56 y +15 z −2 w +32)>>6;
f 2( x, y, z, w )=(−5 x +37 y +37 z −5 w +32)>>6; and
f 3( x, y, z, w )=(−2 x +15 y +56 z −5 w +32)>>6.
13. The computer-readable storage medium according to claim 8 , comprising code for interpolating the at least one aligned sub-pixel value utilizing at least one of a plurality of polyphase filters to filter the integer pixel values.
14. The computer-readable storage medium according to claim 13 , comprising code for interpolating the at least one non-aligned sub-pixel value utilizing at least one of a plurality of polyphase filters to filter the interpolated at least one aligned sub-pixel value.
15. A system for processing a video signal, the system comprising:
A computer comprising: an interpolator, said interpolator interpolates pixels utilizing at most two passes, wherein,
during a first of the two passes,
the interpolator interpolates at least one aligned sub-pixel value utilizing a plurality of integer pixel values, if the at least one aligned sub-pixel value is horizontally or vertically aligned with the plurality of integer pixel values; and
during a second of the two passes,
the interpolator interpolates at least one non-aligned sub-pixel value utilizing the interpolated at least one aligned sub-pixel value, if the at least one non-aligned sub-pixel value is not horizontally or vertically aligned with the plurality of integer pixel values.
16. The system according to claim 15 , comprising a 4-tap filter for filtering the integer pixel values during interpolation of the at least one aligned sub-pixel value.
17. The system according to claim 15 , comprising a 4-tap filter for filtering the integer pixel values during interpolation of the at least one aligned sub-pixel value, wherein the 4-tap filter utilizes at least one of the following filtering operations:
f 1( x, y, z, w )=(−5 x +56 y +15 z −2 w +32)>>6;
f 2( x, y, z, w )=(−5 x +37 y +37 z −5 w +32)>>6; and
f 3( x, y, z, w )=(−2 x +15 y +56 z −5 w +32)>>6.
18. The system according to claim 15 , comprising a 4-tap filter for filtering the interpolated at least one aligned sub-pixel value during interpolation of the at least one non-aligned sub-pixel value.
19. The system according to claim 15 , comprising a 4-tap filter for filtering the interpolated at least one aligned sub-pixel value during interpolation of the at least one non-aligned sub-pixel value, wherein the 4-tap filter utilizes at least one of the following filtering operations:
f 1( x, y, z, w )=(−5 x +56 y +15 z −2 w +32)>>6;
f 2( x, y, z, w )=(−5 x +37 y +37 z −5 w +32)>>6; and
f 3( x, y, z, w )=(−2 x +15 y +56 z −5 w +32)>>6.
20. The system according to claim 15 , wherein the interpolator interpolates the at least one aligned sub-pixel value utilizing at least one of a plurality of polyphase filters to filter the integer pixel values.
21. The system according to claim 20 , wherein the interpolator interpolates the at least one non-aligned sub-pixel value utilizing at least one of a plurality of polyphase filters to filter the interpolated at least one aligned sub-pixel value.
22. A method for processing a video signal, the method comprising: using processor to perform the following steps:
receiving video encoded in accordance with a digital video compression format; and
interpolating pixels of the video utilizing at most two passes of filtering wherein,
during a first pass of the two passes,
interpolating at least one aligned sub-pixel value utilizing a plurality of integer pixel values, if the at least one aligned sub-pixel value is horizontally or vertically aligned with the plurality of integer pixel values; and
during a second of the two passes,
interpolating at least one non-aligned sub-pixel value utilizing the interpolated at least one aligned sub-pixel value, if the at least one non-aligned sub-pixel value is not horizontally or vertically aligned with the plurality of integer pixel values.
23. A system for processing a video signal comprising:
A computer comprising: an interpolator, said interpolator interpolates pixels of digital video data utilizing at most two passes of filtering, wherein,
during a first of the two passes,
the interpolator interpolates at least one aligned sub-pixel value utilizing a plurality of integer pixel values, if the at least one aligned sub-pixel value is horizontally or vertically aligned with the plurality of integer pixel values; and
during a second of the two passes,
the interpolator interpolates at least one non-aligned sub-pixel value utilizing the interpolated at least one aligned sub-pixel value, if the at least one non-aligned sub-pixel value is not horizontally or vertically aligned with the plurality of integer pixel values.