Motion adaptive directional deinterlacing in a video decoder

1. A method for interpolating a pixel for a pixel position, said method comprising:

receiving a plurality of pixels above the pixel position in circuit;

receiving a plurality of pixels below the pixel position;

measuring differences of a plurality pixel pairs from the plurality pixels above the pixel position and the plurality of pixels below the pixel position, each of said plurality of pixel pairs defining a respective one of a plurality of directions between the pixels above the pixel position and the pixels below the pixel position;

selecting one of the plurality of pixel pairs defining a particular one of the plurality of directions, wherein the one of the plurality of pixel pairs has the lowest difference; and

interpolating a spatially interpolated pixel based at least in part on one of the plurality of pixels above the pixel position and below the pixel position, wherein at least one of the pixels above the pixel position are at least three rows above the pixel position, and wherein at least one of the pixels below the pixel position are at least three rows below the pixel position.

2. The method of claim 1 , further comprising:

measuring the motion between a field and a complimentary field, wherein the field comprises the pixels above the pixel position and the pixels below the pixel position.

3. The method of claim 2 , further comprising:

calculating a weighted average from at least one of the plurality of directions and a horizontal direction, thereby generating a spatially interpolated pixel; and

calculating a weighted average from the spatially interpolated pixel and a pixel in the complimentary field.

4. The method of claim 3 , wherein the weighted average is a function of the measured motion between the field and the complimentary field.

5. A circuit for interpolating a pixel for a pixel position, said circuit comprising:

a processor; and

memory connected to the processor, said memory storing a plurality of instructions executable by the processor, wherein execution of the instructions causes:

receiving a plurality of pixels above the pixel position;

receiving a plurality of pixels below the pixel position;

approximating gradients in a plurality of directions between the pixels above the pixel position and the pixels below the pixel position; and

determining a median of at least one of the plurality of pixels above the pixel position, at least one of the plurality of pixels below the pixel position and a pixel in the pixel position of a complimentary field;

interpolating the pixel as a function of any pixel of the plurality of pixels above the pixel position, the plurality of pixels below the pixel position, and the pixel in the complimentary field in the pixel position; and

wherein at least one of the pixels above the pixel position are at least three rows above the pixel position, and wherein at least one of the pixels below the pixel position are at least three rows below the pixel position.

6. The circuit of claim 5 , wherein execution of the instructions by the processor further causes:

calculating a weighted average from at least one of the plurality of directions and a horizontal direction, thereby generating a spatially interpolated pixel.

7. The circuit of claim 5 , wherein execution of the instructions by the processor further causes:

measuring the motion between a field and a complimentary field, wherein the field comprises the pixels above the pixel position and the pixels below the pixel position.

8. The circuit of claim 7 , wherein execution of the instructions by the processor further causes:

calculating a weighted average from at least one of the plurality of directions and a horizontal direction, thereby generating a spatially interpolated pixel;

calculating a weighted average from the spatially interpolated pixel and a pixel in the complimentary field.

9. The circuit of claim 8 , wherein the weighted average is a function of the measured motion between the field and the complimentary field.

10. The circuit of claim 1 , wherein interpolating the pixel for the pixel position is always a function of the average of a pixel directly above the pixel position and the below the pixel position.

11. A circuit for interpolating a pixel for a pixel position, said circuit comprising:

a processor; and

memory connected to the processor, said memory storing a plurality of instructions executable by the processor, wherein execution of the instructions causes:

receiving a plurality of pixels above the pixel position;

receiving a plurality of pixels below the pixel position;

approximating gradients in a plurality of directions between the pixels above the pixel position and the pixels below the pixel position; and

determining a median of at least one of the plurality of pixels above the pixel position, at least one of the plurality of pixels below the pixel position and a pixel in the pixel position of a complimentary field;

interpolating the pixel as a function of any pixel of the plurality of pixels above the pixel position, the plurality of pixels below the pixel position, and the pixel in the complimentary field in the pixel position; and

calculating a weighted average from at least one of the plurality of directions and a horizontal direction, thereby generating a spatially interpolated pixel.

12. The circuit of claim 11 , wherein execution of the instructions by the processor further causes:

measuring the motion between a field and a complimentary field, wherein the field comprises the pixels above the pixel position and the pixels below the pixel position.

13. The circuit of claim 12 , wherein execution of the instructions by the processor further causes:

calculating a weighted average from at least one of the plurality of directions and a horizontal direction, thereby generating a spatially interpolated pixel;

calculating a weighted average from the spatially interpolated pixel and a pixel in the complimentary field.

14. The circuit of claim 13 , wherein the weighted average is a function of the measured motion between the field and the complimentary field.

15. The circuit 11 , wherein approximating gradients in a plurality of directions between the pixels above the pixel position and the pixels below the pixel position further comprises calculating an arctangent function of a gradient in a vertical direction and a gradient in the horizontal direction.

16. The method of claim 3 , wherein the horizontal direction does not intersect the pixel position.

17. The circuit of claim 11 , wherein the horizontal direction does not intersect the pixel position.