Granted Patent B2

US 9,025,903 · App. 13/221,950 · Granted May 5, 2015

Abstract

Claims (57)

1. An image processing device, comprising:

a first edge strength calculation part that calculates a first edge strength for a focus pixel based on pixel values in a first region that includes the focus pixel in an input image;

a second edge strength calculation part that calculates a second edge strength for the focus pixel based on pixel values in a second region that is smaller than the first region and that includes the focus pixel;

a flatness calculation part

that calculates a plurality of first flatnesses for a plurality of fourth regions within a third region, first flatness increasing as variation of pixel values within a respective fourth region decreases, the third region including the focus pixel and pixels neighboring the focus pixel, and at least one of the plurality of fourth regions not including the focus pixel, and

that selects as a second flatness a maximum first flatness from among the plurality of first flatnesses, the second flatness indicating a most flat fourth region from among the plurality of fourth regions; and

a filter determination part that determines a filter coefficient, so that strength of the filter

becomes stronger as the first edge strength becomes larger, the second edge strength becomes smaller, and the second flatness becomes larger, and

becomes weaker as the first edge strength becomes smaller, the second edges strength becomes larger, and the second flatness becomes smaller; and

a filter processing part that filters the input image using the determined filter coefficient and obtains filtered pixel values of the focus pixel.

2. The image processing device of claim 1 , wherein

the filter determination part determines the filter coefficient such that a smoothing strength of the filter becomes higher as a difference value between the first edge strength and the second edge strength increases.

3. The image processing device of claim 2 , wherein

the first edge strength is a difference value between a maximum value and a minimum value of the pixel values included in the first region, and

the second edge strength is a difference value between a maximum value and a minimum value of the pixel values included in the second region.

4. The image processing device of claim 2 , wherein

the first edge strength is one of:

a maximum value of first-order derivative values respectively calculated for pixels in the first region,

a maximum value of second-order derivative values respectively calculated for pixels in the first region,

a difference between a maximum value and a minimum value of luminance values respectively obtained at the pixels in the first region, and

a maximum value of luminance gradients of the luminance values respectively obtained at the pixels in the first region, and

the second edge strength is one of:

a maximum value of first-order derivative values respectively calculated for pixels in the second region,

a maximum value of second-order derivative values respectively calculated for pixels in the second region,

a first-order derivative value for the focus pixel,

a second-order derivative value for the focus pixel,

a difference between a maximum value and a minimum value of luminance values respectively obtained at the pixels in the second region, and

a value of a luminance gradient of the focus pixel.

5. The image processing device of claim 1 , wherein

the filter is an ε filter, and

the strength of the filter is a threshold value ε of the ε filter.

6. The image processing device of claim 1 , wherein

the filter is a Gaussian filter, and

the strength of the filter is a standard deviation a of the Gaussian filter.

7. The image processing device of claim 1 , wherein

the filter is a bilateral filter, and

the strength of the filter is a standard deviation σ d in a domain direction and a range direction standard deviation σ r in a range direction of the bilateral filter.

8. An image processing method, comprising:

calculating a first edge strength for a focus pixel based on values of pixels in a first region that includes the focus pixel in an input image;

calculating a second edge strength for the focus pixel based on values of pixels in a second region that is smaller than the first region and that includes the focus pixel;

calculating a plurality of first flatnesses for a plurality of fourth regions within a third region, first flatness increasing as variation of pixel values within a respective fourth region decreases, the third region including the focus pixel and pixels neighboring the focus pixel, and at least one of the plurality of fourth regions not including the focus pixel, and

selecting as a second flatness a maximum first flatness from among the plurality of first flatnesses, the second flatness indicating a most flat fourth region from among the plurality of fourth regions; and

determines a filter coefficient, so that strength of the filter

becomes stronger as the first edge strength becomes larger, the second edge strength becomes smaller, and the second flatness becomes larger, and

becomes weaker as the first edge strength becomes smaller, the second edges strength becomes larger, and the second flatness becomes smaller; and

filters the input image using the determined filter coefficient and obtaining filtered pixel values of the focus pixel.

9. The image processing device of claim 8 , wherein

the filter is an ε filter, and

the strength of the filter is calculated as a threshold value ε of the ε filter.

10. The image processing method of claim 8 , wherein

the filter is a Gaussian filter, and

the strength of the filter is calculated as a standard deviation ε of the Gaussian filter.

11. The image processing method of claim 8 , wherein

the filter is a bilateral filter, and

the strength of the filter is calculated as a standard deviation ε d in a domain direction and a range direction standard deviation ε r in a range direction of the bilateral filter.

12. The device according to claim 1 , wherein

the first flatness is a number of pixels, at which absolute values of first order derivative values in x direction and y direction is less than or equal to a threshold.

Assignments (6)