Processing method for image interpolation

• US 8,693,783 B2
• Filed: 11/30/2011
• Issued: 04/08/2014
• Est. Priority Date: 09/29/2011
• Status: Active Grant

• Current Assignee: Altek Corporation
• Original Assignee: Altek Corporation
• Inventors: Huang, Yu-Chen, Wu, Chung-Ta
• Primary Examiner(s): Koziol, Stephen R
• Assistant Examiner(s): Lee, Jonathan S
• Application Number:
  US13/307,002
• Publication Number:
  US 20130084014A1
• Time in Patent Office: 860 Days
• Field of Search: None
• US Class Current: 382/199
• CPC Class Codes: G06T 3/403 Edge-driven scaling; Edge-based scaling

Abstract

A processing method for image interpolation is provided. In the method, a contour containing a destination pixel and a plurality of contour pixels is produced and detected by using an edge detection method for a selected area in an original image. Next, two extreme points are identified from the contour pixels by judging the changing trend of brightness value of the contour pixels on the contour. Then using the changing trend and two extreme points determines whether the destination pixel is located at a transition area or a non-transition area. If the destination pixel is located at the non-transition area, an interpolation adjust value is calculated according to the brightness values of the two extreme points and the brightness value of the destination pixel. An interpolation result of the destination pixel is obtained by performing the interpolation on the selected area according to the interpolation adjust value.

Claims

1. A processing method for image interpolation, comprising:

detecting whether an edge exists for a selected area in an original image by using an edge detection method, and determining an edge direction of the edge;

on a normal direction to the edge direction, producing a contour by using a plurality of pixels in the original image to interpolate a plurality of contour pixels, wherein a destination pixel is located on the contour;

identifying two extreme points from the contour pixels by judging a changing trend of a brightness value of the contour;

determining whether the destination pixel is located at a transition area or a non-transition area;

when the destination pixel is located at the non-transition area, calculating an interpolation adjust value according to the brightness values of the two extreme points and the brightness value of the destination pixel; and

obtaining an interpolation result of the destination pixel by performing the interpolation on the selected area according to the interpolation adjust value,

wherein the step of determining whether the destination pixel is located at the transition area or the non-transition area comprises;

according to the brightness values of the two extreme points and the brightness value of the destination pixel, determining the brightness value of the destination pixel is near the brightness value of one extreme point of the two extreme points, and defining the closer extreme point as a destination extreme point;

determining whether an inflexion point exists among the contour pixels from the location of the destination pixel toward an end of the destination extreme point;

when the inflexion point exists among the contour pixels, the destination pixel is located at the non-transition area; and

when the inflexion point does not exist among the contour pixels, the destination pixel is located at the transition area.

2. The processing method for image interpolation as claimed in claim 1, wherein the step of judging the changing trend of the brightness value of the contour comprises:

determining whether the changing trend of the brightness value of the contour is an increasing trend or a decreasing trend by using the plurality of contour pixels near the destination pixel on the contour.

3. The processing method for image interpolation as claimed in claim 2, wherein:

the two extreme points comprise a maximum extreme point and a minimum extreme point,

wherein the maximum extreme point is the contour pixel on the contour having the highest brightness value on the increasing trend direction from location of the destination pixel, and the minimum extreme point is the contour pixel on the contour having the lowest brightness value on the decreasing trend direction from location of the destination pixel.

4. The processing method for image interpolation as claimed in claim 1, wherein the step of determining whether the inflexion point exists among the contour pixels from the location of the destination pixel toward the end of the destination extreme point comprises:

when the destination extreme point is located at the left side of the contour, if the changing trend of the brightness value of the contour pixels at the left side of the contour is opposite from the changing trend of the brightness value of the contour, or the changing trend of the brightness value of the plurality of pixels starting from the left side is lower than a threshold value, the inflexion point exists, and conversely the inflexion point does not exist; and

when the destination extreme point is located at the right side of the contour, if the changing trend of the brightness value of the contour pixels at the right side of the contour is opposite from the changing trend of the brightness value of the contour, or the changing trend of the brightness value of the plurality of pixels starting from the right side is lower than a threshold value, the inflexion point exists, and conversely the inflexion point does not exist.

5. The processing method for image interpolation as claimed in claim 1, wherein the step of calculating the interpolation adjust value according to the brightness values of the two extreme points and the brightness value of the destination pixel comprises:

calculating a center point of the maximum extreme point and the minimum extreme point, and obtaining the brightness value of the center point by calculating an arithmetic average of the brightness value of the maximum extreme point and the brightness value of the minimum extreme point;

when the brightness value of the destination pixel is higher than the brightness value of the center point, the destination extreme point is the maximum extreme point, and conversely, the destination extreme point is the minimum extreme point; and

calculating the interpolation adjust value by using the brightness value of the center point, the brightness value of the destination pixel, and the brightness value of the destination extreme point.

6. The processing method for image interpolation as claimed in claim 1, wherein the step of obtaining the interpolation result of the destination pixel by performing the interpolation on the selected area according to the interpolation adjust value comprises:

performing a bilateral interpolation on the selected area according to the interpolation adjust value to obtain the interpolation result of the destination pixel.

7. The processing method for image interpolation as claimed in claim 1, further comprising:

when the destination pixel is located at the transition area, performing a bilinear interpolation on the selected area to obtain an interpolation result of the destination pixel.

8. The processing method for image interpolation as claimed in claim 1, wherein before the step of determining whether the destination pixel is located at the transition area or the non-transition area, the processing method further comprises:

detecting the edge strength and the edge direction of all the pixels in the selected area by using the edge detection method, so as to judge an area type of the destination pixel, wherein the area type comprises an edge area, a smooth area, and a texture area.

9. The processing method for image interpolation as claimed in claim 8, wherein:

when the area type of the destination pixel is the edge area, subsequently determining whether the destination pixel is located at the transition area or the non-transition area,

when the area type of the destination pixel is the smooth area, performing the bilinear interpolation on the selected area,

and when the area type of the destination pixel is the texture area, performing the bicubic interpolation on the selected area.