Robust automatic page size detection algorithm for scan application

• US 7,545,535 B2
• Filed: 12/21/2005
• Issued: 06/09/2009
• Est. Priority Date: 12/21/2005
• Status: Expired due to Fees

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Zhen, He
• Primary Examiner(s): LEE, TOMMY D
• Application Number:
  US11/312,425
• Publication Number:
  US 20070146811A1
• Time in Patent Office: 1,266 Days
• Field of Search: 358/1.2, 358/1.9, 358/449, 358/464, 358/474
• US Class Current: 358/1.9
• CPC Class Codes: H04N 1/00681 Detecting the presence, position or size of a sheet or correcting its position before scanning H04N1/047 takes precedence; mode signalling H04N1/333; skew detection or correction in image signals H04N1/3878
• CPC Class Codes: H04N 1/0071 Width
• CPC Class Codes: H04N 1/00713 Length
• CPC Class Codes: H04N 1/00737 using the scanning elements as detectors
• CPC Class Codes: H04N 1/00758 using a prescan
• CPC Class Codes: H04N 1/00779 Adjusting settings, e.g. mode, feeding rate or type of paper
• CPC Class Codes: H04N 1/38 Circuits or arrangements for blanking or otherwise eliminating unwanted parts of pictures H04N1/387 takes precedence

Abstract

The present invention discloses a method of automatically detecting a page size for a digital scanning operation. The method includes establishing vertical and horizontal limits of an image area containing image content features of a pre-scanned image, detecting vertical page edge information of the image area, and detecting horizontal page edge information of the image area, such that the vertical and horizontal page edge information define the page size for the pre-scanned image.

Claims

1. A method of automatically detecting a page size for a digital scanning operation, the method comprising:

establishing vertical and horizontal limits of an image area containing image content features of a pre-scanned image by analyzing edge and gray level attributes of the pre-scanned image, wherein establishing of the image area limits comprises;

apportioning the pre-scanned image into a plurality of overlapping localized areas,

detecting edge and gray level attributes for each overlapping localized area, and

qualifying the detected edge and/or gray level attributes to generate the image area;

detecting vertical page edge information of the image area; and

detecting horizontal page edge information of the image area,

wherein the vertical and horizontal page edge information define the page size for the pre-scanned image.

2. The method of claim 1, wherein the horizontal page edge information detection comprises:

filtering the image area along a vertical direction to produce a smoothed image area,

detecting vertical edges in the smoothed image area to generate a vertical edge point map,

dividing the vertical edge point map into overlapping vertical bands; and

determining whether the edge points in the bands exceed a threshold,

wherein the horizontal page edge information is set to the first band from the right that exceeds the threshold and the horizontal page edge information is set to the horizontal limit of the image area when no band exceeds the threshold.

3. A method of automatically detecting a page size for a digital scanning operation, the method comprising:

establishing vertical and horizontal limits of an image area containing image content features of a pre-scanned image by analyzing edge and gray level attributes of the pre-scanned image, wherein establishing of the image area limits comprises;

apportioning the pre-scanned image into a plurality of localized areas,

detecting edge and gray level attributes for each localized area, and

qualifying the detected edge and/or gray level attributes to generate the image area;

detecting vertical page edge information of the image area; and

detecting horizontal page edge information of the image area,

wherein the vertical and horizontal page edge information define the page size for the pre-scanned image,

wherein the qualification of the detected edge and/or gray level attributes comprises qualifying an attribute as an image feature corresponding to a particular localized area if at least another localized area adjacent to the particular localized area also includes a qualified image feature.

4. The method of claim 3, wherein the pre-scanned image is apportioned into non-overlapping localized areas.

5. A method of automatically detecting a page size for a digital scanning operation, the method comprising:

establishing vertical and horizontal limits of an image area containing image content features of a pre-scanned image by analyzing edge and gray level attributes of the pre-scanned image;

detecting vertical page edge information of the image area;

detecting horizontal page edge information of the image area,

wherein the vertical and horizontal page edge information define the page size for the pre-scanned image;

filtering the image area along a horizontal direction to produce a smoothed image area,

detecting horizontal edges in the smoothed image area to generate a horizontal edge point map,

dividing the horizontal edge point map into overlapping horizontal bands; and

determining whether the edge points in the bands exceed a threshold,

wherein the vertical page edge information is set to the first band from the bottom that exceeds the threshold and the vertical page edge information is set to the vertical limit of the image area when no band exceeds the threshold.

6. The method of claim 5, wherein the vertical page edge information detection further comprises detecting whether the page is a full width page and setting a full width flag to circumvent the horizontal page edge information detection.

7. A method of scanning a document comprising:

generating a reduced resolution image of the document;

establishing vertical and horizontal limits of an image area containing image content features of the reduced resolution image by analyzing edge and gray level attributes, wherein establishing of the image area limits comprises;

apportioning the reduced resolution image into a plurality of overlapping localized areas,

detecting edge and gray level attributes for each overlapping localized area, and

qualifying the detected edge and/or gray level attributes to generate the image area;

detecting vertical page edge information of the image area;

detecting horizontal page edge information of the image area;

scanning the document with a corresponding page size based on the detected vertical and horizontal page edge information.

8. The method of claim 7, wherein the horizontal page edge information detection comprises:

filtering the image area along a vertical direction to produce a smoothed image area,

detecting vertical edges in the smoothed image area to generate a vertical edge point map,

dividing the vertical edge point map into overlapping vertical bands; and

determining whether the edge points in the bands exceed a threshold,

wherein the horizontal page edge information is set to the first band from the right that exceeds the threshold and the horizontal page edge information is set to the horizontal limit of the image area when no band exceeds the threshold.

9. A method of scanning a document, comprising:

generating a reduced resolution image of the document;

establishing vertical and horizontal limits of an image area containing image content features of the reduced resolution image by analyzing edge and gray level attributes, wherein establishing of the image area limits comprises;

apportioning the reduced resolution image into a plurality of localized areas,

detecting edge and gray level attributes for each localized area, and

qualifying the detected edge and/or gray level attributes to generate the image area;

detecting vertical page edge information of the image area;

detecting horizontal page edge information of the image area;

scanning the document with a corresponding page size based on the detected vertical and horizontal page edge information,

wherein the qualification of the detected edge and/or gray level attributes comprises qualifying an attribute as an image feature corresponding to a particular localized area if at least another localized area adjacent to the particular localized area also includes a qualified image feature.

10. The method of claim 9, wherein the reduced resolution image is apportioned into non-overlapping localized areas.

11. A method of scanning a document, comprising:

generating a reduced resolution image of the document;

establishing vertical and horizontal limits of an image area containing image content features of the reduced resolution image by analyzing edge and gray level attributes;

detecting vertical page edge information of the image area;

detecting horizontal page edge information of the image area;

scanning the document with a corresponding page size based on the detected vertical and horizontal page edge information;

filtering the image area along a horizontal direction to produce a smoothed image area,

detecting horizontal edges in the smoothed image area to generate a horizontal edge point map,

dividing the horizontal edge point map into overlapping horizontal bands; and

determining whether the edge points in the bands exceed a threshold,

wherein the vertical page edge information is set to the first band from the bottom that exceeds the threshold and the vertical page edge information is set to the vertical limit of the image area when no band exceeds the threshold.

12. The method of claim 11, wherein the vertical page edge information detection further comprises detecting whether the page is a full width page and setting a full width flag to circumvent the horizontal page edge information detection.