Spatial and temporal de-interlacing with error criterion
View Patent ↗A de-interlacing device and method are provided that may be used in a memory based video processor. The de-interlacer mixes the output of a temporal de-interlacer and a spatial de-interlacer. Two separate error values are used; one for the temporal de-interlacer and another for the spatial de-interlacer. The de-interlacing device calculates from the two error values, using a non-linear mapping, a mix factor used to mix between the outputs of the spatial and temporal de-interlacers.
1. A de-interlacer device comprising:
a memory configured to receive and store picture information;
a temporal de-interlacer configured to de-interlace picture information from the memory;
a spatial de-interlacer configured to de-interlace picture information, the spatial deinterlacer configured to calculate a reliability of a directional interpolation;
wherein the reliability of the directional interpolation is used to enable selection between a directional and a vertical interpolation, and wherein the reliability of the directional interpolation is calculated by calculating a luminance gradient approximately along the edge and calculating a luminance gradient approximately perpendicular to the edge and the dominant edge gradient is determined on a macro then a micro level; and
a mixer configured to mix outputs of the temporal de-interlacer and the spatial de-interlacer according to a calculated mix factor;
wherein the calculated mix factor uses at least two error criteria.
2. The de-interlacer device of claim 1 , wherein said the calculated mix factor is a nonlinear calculation.
3. The de-interlacer device of claim 1 , wherein the temporal de-interlacer is configured to calculate a reliability of a temporal interpolation.
4. The de-interlacer of claim 1 , wherein the spatial de-interlacer is configured to calculate a reliability of the directional interpolation.
5. The de-interlacer of claim 1 , wherein the calculated mix factor depends on an expected contribution of outputs of the spatial de-interlacer and the temporal de-interlacer to high frequencies in the picture information.
6. The de-interlacer of claim 1 , wherein the picture information is a video signal.
7. The de-interlacer of claim 1 , wherein the calculated mix factor is based on at least two error functions.
8. The de-interlacer of claim 1 , wherein the at least two error functions are a temporal error function and a spatial error function.
9. The de-interlacer of claim 8 , wherein the spatial error function is a directional error function.
10. The de-interlacer of claim 1 , wherein the spatial de-interlacer and the temporal de interlacer are configured to process the picture information in parallel.
11. A method of de-interlacing a video signal, the method comprising:
de-interlacing the video signal in a temporal de-interlacer;
de-interlacing the video signal in a spatial de-interlacer, the de-interlacing being performed in parallel;
calculating a first error factor and a second error factor;
non-linearly mapping the first error factor and the second error factor to at least one mix factor; and
using the at least one mix factor to mix an output of the temporal de-interlacer and the output of the spatial de-interlacer, wherein the spatial de-interlacer produces a directional intepolated sample and the reliabili of the directional interpolation is calculated by calculating a luminance gradient approximately along the edge and calculating a luminance gradient approximately perpendicular to the edge and a dominant edge gradient is determined on a macro then a micro level.
wherein the mix factor depends on an expected contribution of outputs of the spatial deinterlacer and the temporal de-interlacer to high frequencies in the video signal.
12. The method of claim 11 , wherein the method is used in a memory-based video processor.
13. The method of claim 11 , wherein the first error factor is a temporal error and the second error factor is a spatial error.
14. A memory-based video processor comprising:
a temporal de-interlacer configured to process a video signal;
a spatial de-interlacer configured to process the video signal in parallel with the temporal de-interlacer, wherein the spatial de-interlacer produces a directional interpolated sample and the reliability of the directional interpolation is calculated by calculating a luminance gradient approximately along the edge and calculating a luminance gradient approximately perpendicular to the edge and a dominant edge gradient is determined on a macro then a micro level; and
a mixer configured to receive outputs from the temporal deinterlacer and the spatial de interlaces;
the mixer configured to mix the received outputs according to a non-linear mix factor;
wherein the non-linear mix factor depends on at least two error functions and expected contributions of the spatial and temporal de-interlacer outputs to high vertical frequencies in the video signal.
15. The memory-based video processor of claim 14 , further comprising a memory for storing the video signal.
16. The memory based video processor of claim 14 , wherein the non-linear mix factor is based on a temporal error factor and a directional error factor provided by at least one of the temporal de-interlacer and the spatial de-interlacer.
17. The memory-based video processor of claim 14 wherein the spatial deinterlacer and the temporal de-interlacer are configured to process the picture information in parallel.
18. The memory-based video processor of claim 14 , wherein the spatial deinterlacer is configured to calculate a reliability of a directional interpolation.
19. The memory-based video processor of claim 18 wherein the reliability of the directional interpolation is calculated by calculating a luminance gradient approximately along an edge and calculating a luminance gradient approximately perpendicular to the edge.