Granted Patent B2

US 7,391,476 · App. 11/098,587 · Granted Jun 24, 2008

Abstract

Claims (28)

1. A method of interpolating a pixel of an interline of a first field in a sequence of interlaced fields, the method comprising the steps of:

selecting at least one first pixel, with which a first image information value is associated, from a scan line lying adjacent to the interline, and at least one second pixel, with which a second image information value is associated, from a second scan line lying adjacent to the interline;

providing a third image information value which is dependent on an image information value of at least one third pixel from a second field temporally preceding or following the first field, an image of this second field corresponding to an image position of the pixel to be interpolated, providing a fourth image information value which is dependent on an image information value of at least one fourth pixel which lies vertically adjacent to the third pixel, and providing a fifth image information value which is dependent on an image information value of at least one fifth pixel which lies vertically adjacent to the third pixel;

generating a first filter value by low-pass filtering the at least one first and second image information values;

determining a width of an interval established by the first and second image information values;

generating a second filter value by high-pass filtering the third, fourth, and fifth image information values;

generating an image information value for the interpolated pixel as a sum of the first filter value and the second filter value when the absolute value of the second filter value is smaller than half the interval width; and

generating an image information value for the interpolated pixel such that the image information value lies within the interval width when the second filter value is greater than half the interval width.

2. The method of claim 1 , where the step of low-pass filtering comprises the step of averaging the first and second image information values.

3. The method of claim 1 , where the step of high-pass filtering comprises the step of averaging the at least one fourth and fifth image information values and generating a difference between the third image information value and the averaged fourth and fifth information values.

4. The method of claim 1 , where the step of generating a second filter value comprises applying a gain factor.

5. The method of claim 1 , where the step of high-pass filtering utilizes a high-pass filter with the following response:

y=x+a for x<−a

y=0 for − a≦x≦a

y=x−a for x>a

where x is the filter input signal, y is the filter output signal, and a>0.

6. The method of claim 1 , where the third image information value corresponds to the image information value of the third pixel, the fourth image information value corresponds to the image information value of the fourth pixel, and the fifth image information value corresponds to the image information value of the fifth pixel.

7. The method of claim 1 , where the third image information value is generated by linear averaging of the image information value of the third pixel as well as from image information values of additional pixels at the pixel position of the third pixel from temporally preceding or following fields, where the fourth image information value is generated by linear averaging of the image information value of the fourth pixel as well as from image information values of additional pixels at the pixel position of the fourth pixel from temporally preceding or following fields, and where the fifth image information value is generated by linear averaging of the image information value of the fifth pixel as well as from image information values of additional pixels at the pixel position of the fifth pixel from temporally preceding or following fields.

8. The method according to claim 1 , further comprising the step of outputting the first or second image information value as the image information value of the interpolated pixel when the second filter value is greater than half the interval width.

9. The method of claim 1 , further comprising the step of mixing the first and second image information values to generate the image information value of the interpolated pixel when the second filter value is greater than half the interval width.

10. A pixel interpolation device, comprising:

a first filter device that comprises a low-pass filter and receives first and second image information values and which provides a first filter signal value;

a second filter device that comprises a high-pass filter and receives third, fourth, and fifth image information values, and which provides a second filter signal value;

an interval determination unit that receives the first and second image information values and which provides an interval value which is dependent on a width of an interval spanned by the first and second image information values; and

a signal processing unit that receives the first filter signal value, the second filter signal value, and the interval value, and provides an interpolated image information value dependent on a comparison of the interval width with the second filter signal value;

where the signal processing unit provides at its output the sum of the first and second filter values as the image information value when the absolute value of the second filter signal value is less than the interval width.

11. The device of claim 10 , where the signal processing unit provides as the image information value a value which lies within the interval spanned by the first and second image information values when the absolute value of the second filter signal value is greater than the interval width.

12. The device of claim 11 , where the interval width corresponds to 0.5 times the interval width.

Assignments (9)