Method and apparatus for distinguishing between noisy continuous tone document types and other document types to maintain reliable image segmentation

• US 6,411,735 B1
• Filed: 09/23/1998
• Issued: 06/25/2002
• Est. Priority Date: 09/23/1998
• Status: Expired due to Term

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Schweid, Stuart A., Clark, Raymond J., Williams, Leon C.
• Primary Examiner(s): Johns, Andrew W.
• Assistant Examiner(s): AZARIAN, SEYED H
• Application Number:
  US09/158,868
• Time in Patent Office: 1,371 Days
• Field of Search: 382/224, 382/225, 382/226, 382/227, 382/228, 382/229, 382/220, 382/173, 382/176, 382/180, 358/462
• US Class Current: 382/224
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables
• CPC Class Codes: H04N 1/40062 Discrimination between different image types, e.g. two-tone, continuous tone

Abstract

A method and apparatus are disclosed for distinguishing noisy continuous tone document types from other document types such as halftone or text, in order to maintain reliable image segmentation. In both cases a Range of Video microclassifier is added to the other microclassifiers (Sij, estimated halftone frequency, etc.) to disqualify noisy continuous tone images from being incorrectly classified as other document types such as halftone or text document types.

Claims

1. A method for classifying a pixel of image data as one of a plurality of image types, comprising the steps of:

(a) determining a maximum video value within a context of pixels in a vicinity of the pixel being classified;

(b) determining a minimum video value within the context of pixels;

(c) determining a range of video value for the pixel being classified by subtracting the minimum video value from the maximum video value;

(d) determining a conditional document type classification for the pixel being classified;

(e) comparing the range of video value with a predetermined threshold value;

(f) classifying the pixel as a contone document type when the range of video value is less than the threshold value; and

(g) classifying the pixel as the conditional document type when the range of video value is greater than the threshold value.

2. The method as claimed in claim 1, wherein step (c) includes the step of:

(h) determining the range of video value for a context of pixels in a vicinity of the pixel being classified.

3. The method as claimed in claim 1, wherein step (d) includes the step of:

(h) determining the conditional document type classification based on at least an absolute value of the sum of the laplacians for the pixel being classified.

4. The method as claimed in claim (1), wherein step (a) includes the step of:

(h) centering the context of pixels on the pixel being classified.

5. The method as claimed in claim 1, wherein step (a) includes the step of:

(h) determining the range of video value for a 5×

5 context of pixels in a vicinity of the pixel being classified.

6. The method as claimed in claim 1, further including the steps of:

(h) processing the pixel based on the conditional document type classification of the pixel, and

(i) rendering the image processed pixel.

7. The method as claimed in claim 1, further including the steps of:

(h) processing the pixel based on a contone document type classification of the pixel, and

(i) rendering the image processed pixel.

8. A method for classifying a pixel of image data as one of a plurality of image types, comprising the steps of:

(a) determining a maximum video value within a context of pixels in a vicinity of the pixel being classified;

(b) determining a minimum video value within the context of pixels;

(c) determining a range of video value for the pixel being classified by subtracting the minimum video value from the maximum video value;

(d) separating a large number of small video variations from a small number of large video variations;

(e) determining a conditional document type classification for the pixel being classified;

(f) comparing the range of video value with a predetermined threshold value;

(g) classifying the pixel as a contone document type when the range of video value is less than the threshold value; and

(h) classifying the pixel as the conditional document type when the range of video value is greater than the threshold value.

9. The method as claimed in claim 8, wherein the subtracting the minimum video value from the maximum video value is determined by an equation expressed as Rij=V_MAX−

V_MIN.

10. The method as claimed in claim 8, wherein a noisy contone document is differentiated from other document types.

11. A method for classifying a pixel in image signals, comprising the steps of:

(a) determining a maximum video value within a context of pixels in a vicinity of the pixel being classified;

(b) determining a minimum video value within the context of pixels;

(c) determining a range of video value for the pixel being classified by subtracting the minimum video value from the maximum video value;

(d) separating a large number of small video variations from a small number of large video variations;

(e) determining a conditional document type classification for the pixel being classified;

(f) comparing the range of video value with a predetermined threshold value;

(g) classifying the pixel as a contone document type when the range of video value is less than the threshold value;

(h) classifying the pixel as the conditional document type when the range of video value is greater than the threshold value;

(i) processing the pixel based on the image type classification of the pixel, and

(j) rendering the image processed pixel.