Method and device for converting video signals

• US 7,315,331 B2
• Filed: 01/08/2002
• Issued: 01/01/2008
• Est. Priority Date: 01/09/2001
• Status: Expired due to Fees

• Current Assignee: Dynamic Data Technologies LLC
• Original Assignee: Micronas Gmbh (TDK Corporation)
• Inventors: Franzen, Ortwin Ludger
• Primary Examiner(s): HSIA, SHERRIE Y
• Application Number:
  US10/250,704
• Publication Number:
  US 20040119887A1
• Time in Patent Office: 2,184 Days
• Field of Search: 348/452, 348/441, 348/446, 348/448, 348/451, 348/458, 348/459
• US Class Current: 348/452
• CPC Class Codes: H04N 5/145 Movement estimation for video coding H04N19/51
• CPC Class Codes: H04N 7/012 Conversion between an interlaced and a progressive signal for graphics images G09G2310/0229
• CPC Class Codes: H04N 7/0132 the field or frame frequency of the incoming video signal being multiplied by a positive integer, e.g. for flicker reduction
• CPC Class Codes: H04N 7/014 involving the use of motion vectors motion estimation and compensation in video coding H04N19/51

Abstract

A method for the motion-vector-based conversion of a first video signal which contains a first image sequence at a first frequency to a second video signal which contains a second image sequence at a second frequency, where at least some of the images of the second image sequence which are phase-shifted relative to the images of the first image sequence are generated by an interpolation of the images of the first image sequence such that, for a pixel of an image from the second image sequence, pixels at least from one first image and from one second image of the first image sequence are filtered by a median filter, where the median filter is an adaptively weighted median filter.

Claims

1. A method for motion-vector-based conversion of a first video signal having a first image sequence at a first frequency to a second video signal having a second image sequence at a second frequency, the method comprising the steps of:

interpolating at least one intermediate image that is part of the second image sequence, where the interpolated intermediate image lies temporally between first and second images in the first image sequence, where the at least one intermediate image lies at a first temporal distance from the first image in the first image sequence and lies at a second temporal distance from the second image in the first image sequence, where the first and second temporal distances define an interpolation phase, where for a pixel of the interpolated image in the second image sequence a plurality of pixels at least from each of the first and second images of the first image sequence are median filtered; and

adaptively weighting the median filter as a function of a first projection factor that comprises a normalized temporal distance for a period of time between the first image of the first image sequence and the intermediate image and as a function of a second projection factor that comprises a normalized temporal distance for a period of time between the intermediate image and the second image of the first image sequence.

2. The method of claim 1, where the step of adaptively weighting further adaptively weights the median filter as a function of a motion vector assigned to a predetermined at least one pixel.

3. The method of claim 1, the method further comprising the steps of:

generating a first filter mask for each of a plurality of defined interpolation phases by determining the selection and weighting of the plurality of pixels of the first image of the first image sequence;

generating a second filter mask for each of a plurality of defined interpolation phases by determining the selection and weighting of the plurality of pixels of the second image of the first image sequence; and

median filtering using the generated first and second filter masks to generate the at least one intermediate image.

4. The method of claim 3, where each of the first and second filter masks comprise a central pixel that is weighted more strongly than the pixels surrounding the central pixel, where the sum of the weights of the central pixels is greater than the sum of the weights of the remaining pixels surrounding the central pixels.

5. The method of claim 3, where each of the first and second filter masks are such that the larger the number of the pixels processed by the first and the second filter masks from the first image and the second image of the first image sequence the smaller is a corresponding one of the first and second projection factors.

6. The method of claim 5, where the filter masks are such that the majority of the filter weights correspond to the one of the first and the second filter masks assigned to the smallest one of the first and second projection factors.

7. The method of claim 3, where for each defined interpolation phase, more than one first filter mask and more than one second filter mask are provided that each filter certain structured image regions, and where the more than one first and second filter masks are selected as a function of a detection of such image regions.

8. The method of claim 3, where for each defined interpolation phase, more than one first filter mask and more than one second filter mask are provided that each filter in connection with motion vectors of varying reliability, where the more than one first and second filter masks are selected as a function of an estimated reliability of the motion vector.

9. The method of claim 8, further comprising the step of providing pixels of the interpolated intermediate image to the median filter.

10. The method of claim 1, further comprising the step of separating the first video signal into at least two subband signals by one of a linear and a nonlinear band separation.

11. The method of claim 1, where the first video signal comprises an interlaced signal.

12. The method of claim 1, where the first video signal comprises a progressive signal.

13. A device for the motion-vector-based conversion of a first video signal which contains a first image sequence at a first frequency to a second video signal which contains a second image sequence at a second frequency, the device comprising:

means for motion estimation which provides a motion vector;

an interpolation system with a median filter which is designed to generate images of the second image sequence which are phase-shifted relative to images of the first image sequence such that for a pixel of an image from the second image sequence, pixels at least from one first image and one second image of the first image sequence are filtered;

means for providing a projection factor, where the particular projection factor is the temporal distance, normalized for the temporal distance between the first and second image, between the image and the first image, or between the image and the second image; and

a filter mask selection unit coupled to an interpolation system, to which at least the projection factor is supplied as the input signal, and which makes available filter masks dependent on the input signals to the interpolation system.

14. The device of claim 13, comprising a vector projection unit, to which the projection factor and the motion vector are supplied, and which provides a projected motion vector which is supplied to a filter mask selection unit.

15. The device of claim 13, comprising:

means for band separation which provides at least one first band signal and at least one second band signal from the first video signal;

a first interpolation system connected to a first filter mask selection unit, and a second interpolation system connected to a second filter mask selection unit, where the at least one first band signal is supplied to the first interpolation system, and the at least one second band signal is supplied to the second interpolation system; and

a logic circuit, to which an output signal of the first interpolation system and an output signal of the second interpolation system are supplied, and which provides the second video signal from the output signal of the first interpolation system and from the output signal of the second interpolation system.

16. A device for the motion-vector-based conversion of a first video signal which contains a first image sequence at a first frequency to a second video signal which contains a second image sequence at a second frequency, the device comprising:

a motion estimation unit that provides a motion vector;

an interpolation system with a median filter that generates images of the second image sequence that are phase-shifted relative to images of the first image sequence, such that for a pixel of an image from the second image sequence pixels at least from one first image and one second image of the first image sequence are filtered;

means for providing a projection factor, where the particular projection factor is the temporal distance, normalized for the temporal distance between the first and second image, between the image and the first image, or between the image and the second image; and

a filter mask selection unit coupled to an interpolation system, to which at least the projection factor is supplied as the input signal, and which provides filter masks dependent on the input signals to the interpolation system.