IMAGE PROCESSING APPARATUS AND METHOD THEREOF

• US 20100092092A1
• Filed: 08/29/2006
• Published: 04/15/2010
• Est. Priority Date: 08/31/2005
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Sakaue, Tsutomu
• Granted Patent: US 8,004,732 B2
• Time in Patent Office: Days
• US Class Current: 382/200
• CPC Class Codes: H04N 1/40062 Discrimination between different image types, e.g. two-tone, continuous tone

Abstract

When a halftone dot region includes a character or line drawing, high extraction performance of a halftone dot region cannot be obtained. Hence, the isolation amount of an interest pixel is calculated, and the isolation amounts are integrated in a plurality of areas having different sizes. Thresholds set for respective areas are compared with the integration result to determine whether or not the interest pixel belongs to a halftone dot region.

Claims

1-9. -9. (canceled)

10. A method of extracting image regions from an image including different types of image regions, comprising:

using a processor to perform the steps of;

generating an edge-emphasized image by applying an edge-emphasis process to an input image;

setting an inner edge signal of a pixel of the edge-emphasized image to an affirmative value when a value of the pixel greater than a first threshold value, and setting an outer edge signal of a pixel of the edge-emphasized image to an affirmative value when a value of the pixel less than a second threshold value, wherein the first threshold value is positive and the second threshold value is negative;

determining whether an OR signal of each pixel is set to an affirmative value or not based on the inner and outer edge signals of each pixel;

counting a first number of pixels having the affirmative OR signal in an area which includes a pixel of interest and has a first pixel size so as to compare the counted first number with a third threshold value;

counting a second number of pixels having the affirmative OR signal in an area which includes the pixel of interest and has a second pixel size larger than the first pixel size so as to compare the counted second number with a fourth threshold value; and

determining whether or not the pixel of interest belongs to a halftone dot region based on the comparison results in the first and second counting steps.

11. The method according to claim 10, wherein the first determining step comprising:

a first step of setting a first isolation amount determination signal of each pixel to an affirmative value when a pattern represented by the inner edge signals of that pixel and pixels surrounding that pixel matches a first predetermined pattern;

a second step of setting a second isolation amount determination signal of each pixel to an affirmative value when a pattern represented by the outer edge signals of that pixel and pixels surrounding that pixel matches a second predetermined pattern;

a third step of determining that the OR signal of a pixel of interest is the affirmative value when at least one of the first and second isolation amount determination signals of the interest pixel and pixels surrounding the pixel of interest is affirmative.

12. The method according to claim 11, wherein in the first step, if it has three and more pixels having the affirmative inner edge signal in four pixels placed in a center of a 4×

4 pixel area which includes a pixel of interest and pixels surrounding the pixel of interest, and if it has six and more pixels having the negative inner edge signal in eight pixels placed in sides of the 4×

4 area without corners, the first isolation amount determination signal of the pixel of interest is set to the affirmative value.

13. The method according to claim 10, further comprising the step of counting a third number of pixels having the affirmative OR signal in an area which includes the pixel of interest and has a third pixel size larger than the second pixel size so as to compare the counted third number with a fifth threshold value,

wherein in the second determining step, an AND result of the comparison results of the first and second counting steps is computed, an AND result of the comparison results of the second and third counting steps is computed, an OR result of the AND results is computed, and it is determined that the pixel of interest belongs to the halftone dot region when the OR result is affirmative.

14. An image processing apparatus for extracting image regions from an image including different types of image regions, comprising:

a generator configured to generate an edge-emphasized image by applying an edge-emphasis process to an input image;

a setter configured to set an inner edge signal of a pixel of the edge-emphasized image to an affirmative value when a value of the pixel greater than a first threshold value, and to set an outer edge signal of a pixel of the edge-emphasized image to an affirmative value when a value of the pixel less than a second threshold value, wherein the first threshold value is positive and the second threshold value is negative;

a first determiner configured to determine whether an OR signal of each pixel is set to an affirmative value or not based on the inner and outer edge signals of each pixel;

a first comparator configured to count a first number of pixels having the affirmative OR signal in an area which includes a pixel of interest and has a first pixel size so as to compare the counted first number with a third threshold value;

a second comparator configured to count a second number of pixels having the affirmative OR signal in an area which includes the pixel of interest and has a second pixel size larger than the first pixel size so as to compare the counted second number with a fourth threshold value; and

a second determiner configured to determine whether or not the pixel of interest belongs to a halftone dot region based on the comparison results of the first and second comparators.

15. A computer-readable storage medium storing a computer-executable program for causing a computer to perform a method of extracting image regions from an image including different types of image regions, the method comprising the steps of:

generating an edge-emphasized image by applying an edge-emphasis process to an input image;

setting an inner edge signal of a pixel of the edge-emphasized image to an affirmative value when a value of the pixel greater than a first threshold value, and setting an outer edge signal of a pixel of the edge-emphasized image to an affirmative value when a value of the pixel less than a second threshold value, wherein the first threshold value is positive and the second threshold value is negative;

determining whether an OR signal of each pixel is set to an affirmative value or not based on the inner and outer edge signals of each pixel;

counting a first number of pixels having the affirmative OR signal in an area which includes a pixel of interest and has a first pixel size so as to compare the counted first number with a third threshold value;

counting a second number of pixels having the affirmative OR signal in an area which includes the pixel of interest and has a second pixel size larger than the first pixel size so as to compare the counted second number with a fourth threshold value; and

determining whether or not the pixel of interest belongs to a halftone dot region based on the comparison results in the first and second counting steps.