System for detecting an edge of an image

• US 5,444,798 A
• Filed: 12/28/1993
• Issued: 08/22/1995
• Est. Priority Date: 03/18/1991
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Enomoto, Hajime, Miyamura, Isao
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): KELLEY, CHRISTOPHER S
• Application Number:
  US08/174,627
• Time in Patent Office: 602 Days
• Field of Search: 382/22, 382/56, 382/17, 348/396, 348/486
• US Class Current: 382/199
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/13 Edge detection

Abstract

A system for correctly detecting an edge of an image in which a first set of edges are corrected by a second set of edges which are detected as errors between an original image and a reproduced image obtained using the first set of edges. The edges may be in luminance components or lamellar chrominance components of a color image. A third set of edges obtained by applying a smaller threshold to the original image may be used to refine the second set of edges.

Claims

1. A system for detecting an edge of an image, comprising:

first edge detecting means for extracting a first edge of an original image using a first threshold;

reproduction processing means, operatively connected to said first edge detecting means, for obtaining a reproduced image corresponding to the original image using the first edge;

error extracting means, operatively connected to said reproduction processing means, for extracting an error between the original image and the reproduced image;

second edge detecting means, operatively connected to said error extracting means, for detecting a second edge in the error extracted by said error extracting means using a second threshold smaller than the first threshold; and

edge correcting means, operatively connected to said first edge detecting means and said second edge detecting means, for correcting said first edge by logically adding said first edge and said second edge to obtain a corrected edge in a corrected reproduced image.

2. A system for detecting an edge of an image as claimed in claim 1, further comprising:

image separating means for separating the original image into image luminance components and image chrominance components, and

error separating means for separating the error into error luminance components and error chrominance components, the first edge represented by at least one of the image luminance components and the image chrominance components, and the second edge represented by at least one of the error luminance components and the error chrominance components.

3. A system for detecting an edge of an image as claimed in claim 2, wherein the first edge includes image luminance divergence components that exceed said first threshold of the image luminance components, and the second edge includes error luminance divergence components that exceed the second threshold of the error luminance components.

4. A system for detecting an edge of an image as claimed in claim 2, wherein the first edge includes image chrominance divergence components that exceed the first threshold of the image chrominance components, and the second edge includes error chrominance divergence components that exceed the second threshold of the error chrominance components.

5. A system for detecting an edge of an image as claimed in claim 2, further comprising:

image chrominance separating means for separating the image chrominance components into image lamellar components and image vortex components, and

error chrominance separating means for separating the error chrominance components into error lamellar components and error vortex components, the first edge represented by at least one of the image lamellar components and the image vortex components, and the second edge represented by at least one of the error luminance components and the error chrominance components.

6. A system for detecting an edge of an image as claimed in claim 5, wherein the first edge including image lamellar divergence components that exceed the first threshold of the image lamellar components, and the second edge includes error lamellar divergence components that exceed the second threshold of the error lamellar components.

7. A system for detecting an edge of an image as claimed in claim 5, wherein the first edge includes image vortex divergence components that exceed the first threshold of the image vortex components, and the second edge includes error vortex divergence components that exceed the second threshold of the error vortex components.

8. A system for detecting an edge of an image as claimed in claim 1, wherein said error extracting means comprises bulk error detecting means for extracting a bulk error from the error, the bulk error expanding to a certain indicating error in an area.

9. A system for detecting an edge of an image as claimed in claim 1, wherein said bulk error detecting means comprises:

absolute value obtaining means for obtaining the absolute value of a difference between the reproduced image output from said reproduction processing means and the original image,

integration means for calculating an integration of the absolute value output by said absolute value obtaining means over a predetermined area, and

comparing means for comparing the integration output by said integration means with a predetermined threshold.

10. A system for detecting an edge of an image as claimed in claim 1, further comprising:

third edge detecting means, operatively connected to receive the original image and to said edge correcting means, for detecting a third edge of the original image using a third threshold smaller than the first threshold, and

comparing means, operatively connected to said second edge detecting means and to said third edge detecting means, for obtaining a difference between the second edge and the third edge, the first edge being corrected by adding the difference to the first edge.

11. A system for detecting an edge of an image as claimed in claim 1, further comprising image reproducing means, operatively connected to said edge correcting means, for generating the corrected reproduced image using the corrected edge obtained by said edge correcting means.