KR20060089091A - Diagonal interpolator in deinterlacer - Google Patents

Abstract

The present invention relates to a diagonal interpolation method of a deinterlacer. In the related art, since a direction is determined by calculating a difference between pixels connected diagonally, there is a problem of failing to correctly generate a pixel by incorrectly determining a direction and causing a deterioration in image quality. In view of the above problems, the present invention includes the steps of selecting pixels of two upper and lower lines centering on a new pixel; Calculating a pixel difference and an average of the pixel difference for a plurality of paths moving between the pixels using the upper line as the starting point and the lower line as the end point; Selecting a minimum value among average values for a plurality of paths, and extracting a direction correlation by adding a difference between a start point pixel and an end point pixel; Calculating a direction through the shortest path by calculating a direction correlation with respect to another diagonal line and selecting a minimum value among the plurality of direction correlations; Compute interpolation of the new pixels by calculating the weighted average of the pixels located in the shortest path, and interpolate the new pixels by calculating the correlation between pixels to prevent image quality degradation when converting interlaced scan images into progressive scan images It is effective.

Description

Diagonal interpolation method for deinterlacer {DIAGONAL INTERPOLATOR IN DEINTERLACER}

1 is an exemplary diagram illustrating a conventional diagonal interpolation method.

2 is a block diagram showing the configuration of a deinterlacer for explaining the present invention.

3 is a flowchart illustrating a diagonal interpolation method of a deinterlacer according to an embodiment of the present invention.

4 is an exemplary diagram illustrating a diagonal interpolation method of FIG. 3.

The present invention relates to a diagonal interpolation method of a deinterlacer, and more particularly, to a diagonal interpolation method of a deinterlacer that calculates a shortest path between pixels of two lines and weights the pixels of the shortest path to interpolate an image.

NTSC televisions reproduce the screen with interlaced scans, while computer monitors and digital televisions reproduce the screen with progressive scans. An interlaced scan scan divides one frame into two fields to sequentially scan one frame to form one frame, and a progressive scan scans one frame at a time.

De-interlacing is the process of converting an interlaced scan image divided into two fields into a frame before outputting it to the screen. The deinterlacer deinterlaces the interlaced scan image and converts it into a progressive scan image. The deinterlacer interpolates an image by newly generating pixels using a diagonal interpolation method as a deinterlacing method.

1 is an exemplary diagram illustrating a conventional diagonal interpolation method.

The diagonal interpolation method produces a new pixel by obtaining a correlation in a diagonal direction and interpolating in a direction having the maximum correlation. The diagonal interpolation method is effective in creating a new line while preserving the edge in consideration of the local characteristics of the image.

The diagonal interpolation method calculates the difference between pixels connected diagonally on the U line and the L line, selects the direction having the smallest difference, and generates new pixels by interpolating pixels located in the direction.

However, in the prior art as described above, since the direction is determined by calculating the difference between the diagonally connected pixels, the direction is incorrectly determined, thereby failing to generate the correct pixel, resulting in a deterioration in image quality.

Accordingly, the present invention has been made in view of the above-described problems, and provides a diagonal interpolation method of a deinterlacer capable of interpolating an image by calculating a shortest path between pixels of two lines and weighting the pixels of the shortest path. There is a purpose.

The present invention for achieving the above object comprises the steps of selecting the pixels of the upper and lower lines around the new pixel; Calculating a pixel difference and an average of the pixel difference for a plurality of paths moving between the pixels using the upper line as the starting point and the lower line as the end point; Selecting a minimum value among average values for a plurality of paths, and extracting a direction correlation by adding a difference between a start point pixel and an end point pixel; Calculating a direction through the shortest path by calculating a direction correlation with respect to another diagonal line and selecting a minimum value among the plurality of direction correlations; And interpolating a new pixel by calculating a weighted average of pixels located in the shortest path.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 2 is a block diagram showing the configuration of a deinterlacer for explaining the present invention. As shown therein, a pixel input unit 21 for inputting pixels of two upper and lower lines centering on a new pixel; A path selector 22 which determines a start point and an end point of the pixel and selects a moving path; A path value calculator 23 for extracting direction correlation with respect to the selected path; A shortest path selector 24 which determines a shortest path by selecting a minimum value among a plurality of directional correlations; The pixel interpolator 25 calculates a weighted average of pixels with respect to the determined shortest path, determines a new pixel, and interpolates an image.

The pixel input unit 21 inputs two upper and lower pixels with the new pixel as the center, and the path selector 22 moves the pixels between the pixels of the upper line as the starting point and the pixels of the lower line as the end point. Select.

The path value calculator 23 calculates an average of the pixel difference and the pixel difference for the selected path, and extracts the direction correlation by adding the difference between the pixel at the start point and the pixel at the end point.

The shortest path selector 24 instructs the path selector to select a path, receives direction correlations for a plurality of paths, and selects a minimum value among the plurality of direction correlations to determine a direction through the shortest path.

The pixel interpolator 25 calculates a weighted average of pixels using experimentally obtained weights for the determined shortest path, and determines a new pixel to interpolate the image.

3 is an operation flowchart of a diagonal interpolation method of a deinterlacer according to an exemplary embodiment of the present invention, as shown in this example, selecting pixels of two upper and lower lines centering on a new pixel; Calculating a pixel difference and an average of the pixel difference for a plurality of paths moving between the pixels using the upper line as the starting point and the lower line as the end point; Selecting a minimum value among average values for a plurality of paths, and extracting a direction correlation by adding a difference between a start point pixel and an end point pixel; Calculating a direction through the shortest path by calculating a direction correlation with respect to another diagonal line and selecting a minimum value among the plurality of direction correlations; A new pixel is interpolated by calculating a weighted average of pixels located in the shortest path.

The deinterlacer calculates the shortest path in order to select the pixels of the upper and lower lines around the new pixel and extract the correlation between the pixels of the upper line and the pixels of the lower line diagonally.

The deinterlacer calculates an average of the pixel difference and the pixel difference for a plurality of paths moving between the pixels with the upper line as the starting point and the lower line as the end point.

For example, a calculation example of finding the shortest path between U (t-3) and L (t + 3) when extracting the direction correlation between U (t-3) and L (t + 3) from the pixel array of FIG. 4. Explain. When the moving path is Path0, the path value is S (0) = D (U (t-3), U (t-2)) + D (U (t-2), U (t-1)) + D (U (t-1), L (t + 1)) + D (L (t + 1), L (t + 2)) + D (L (t + 2), L (t + 3)) and move When the path is Path1, the path value is S (1) = D (U (t-3), U (t-2)) + D (U (t-2), L (t-1)) + D (L ( t-1), L (t)) + D (L (t), L (t + 1)) + D (L (t + 1), L (t + 2)) + D (L (t + 2) ), L (t + 3)).

The deinterlacer uses Equation 1 to average the pixel difference accumulated value of the path, select the minimum value among a plurality of average values, and add the difference between U (t-3) and L (t + 3) to extract the direction correlation. .

Dist (t-3) = (Min (Ave (S (j))) + D (U (t-3), L (t + 3)))

Here, S (j) is the pixel difference accumulated value of the path, D (U (t-3), L (t + 3)) is the pixel difference, Ave (S (j)) is the average value, and Dist (t- 3) is direction correlation.

The deinterlacer calculates direction correlation with respect to another diagonal line, selects a minimum value among a plurality of direction correlations, determines a direction through the shortest path, and determines a new pixel by calculating a weighted average of pixels located in the shortest path. That is, the deinterlacer calculates a weighted average of pixels using experimentally obtained weights for the determined shortest path, and determines a new pixel to interpolate the image.

As described above in detail, the present invention has an effect of preventing deterioration in image quality when converting an interlaced scan image into a progressive scan image by calculating a correlation between pixels and interpolating new pixels.

Claims (3)

Selecting pixels of two upper and lower lines around the new pixel; Calculating a pixel difference and an average of the pixel difference for a plurality of paths moving between the pixels using the upper line as the starting point and the lower line as the end point; Selecting a minimum value among average values for a plurality of paths, and extracting a direction correlation by adding a difference between a start point pixel and an end point pixel; Calculating a direction through the shortest path by calculating a direction correlation with respect to another diagonal line and selecting a minimum value among the plurality of direction correlations; And interpolating a new pixel by calculating a weighted average of pixels located in the shortest path. The method of claim 1, wherein the direction correlation extraction is performed by averaging pixel difference accumulated values of a path using Equation 1, selecting a minimum value among a plurality of average values, and extracting direction correlation by adding a difference between a start pixel and an end pixel. Diagonal interpolation method of a deinterlacer, characterized in that made. [Equation 1] Dist (t) = (Min (Ave (S (j))) + D (U (start), L (final))) Here, S (j) is the pixel difference accumulated value of the path, D (U (start), L (final)) is the pixel difference, Ave (S (j)) is the average value, and Dist (t) is the direction correlation. to be. The diagonal interpolation method of claim 1, wherein the weighted average of the pixels is obtained by calculating weighted averages of pixels using experimentally obtained weights for the determined shortest paths, and determining new pixels.

Images