Image processing device and method for detecting objects in image data

• US 6,256,411 B1
• Filed: 05/26/1998
• Issued: 07/03/2001
• Est. Priority Date: 05/28/1997
• Status: Expired due to Term

• Current Assignee: Minolta Corporation Limited (Konica Minolta Inc.)
• Original Assignee: Minolta Corporation Limited (Konica Minolta Inc.)
• Inventors: Iida, Kentaro
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): Mariam, Daniel G.
• Application Number:
  US09/084,868
• Time in Patent Office: 1,134 Days
• Field of Search: 382/199, 382/203, 382/317, 382/215, 382/242, 358/453, 358/486, 358/488, 358/491, 358/493, 358/464
• US Class Current: 382/203
• CPC Class Codes: G06T 2207/10008 from scanner, fax or copier
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/70 Determining position or orientation of objects or cameras camera calibration G06T7/80
• CPC Class Codes: H04N 1/00795 Reading arrangements details of scanning heads H04N1/024; scanning arrangements therefor H04N1/04
• CPC Class Codes: H04N 1/1013 with sub-scanning by translatory movement of at least a part of the main-scanning components H04N1/107 takes precedence
• CPC Class Codes: H04N 1/193 using electrically scanned linear arrays , e.g. linear CCD arrays
• CPC Class Codes: H04N 1/38 Circuits or arrangements for blanking or otherwise eliminating unwanted parts of pictures H04N1/387 takes precedence
• CPC Class Codes: H04N 2201/0434 specially adapted for scanning pages of a book
• CPC Class Codes: H04N 2201/0436 Scanning a picture-bearing surface lying face up on a support H04N2201/0434 takes precedence; using cylindrical picture bearing surfaces H04N1/06; with manual scanning H04N1/107

Abstract

An image processing device and method recognizing and deleting extraneous objects in image for data based on analysis of the image data. An object detection means detects an object region such as a finger or the like in a document image, and an image processing means erases part of an image corresponding to the object region of a finger or the like in the image data. The object detection means detects the object region by a dynamic contouring method in which a specific initial contour is set relative to the image data, and the initial contour is adjusted along a border of variable density in the image data so as to accurately erase an object image of a finger of a user holding a document.

Claims

1. An image processing device comprising:

an object detection means for detecting an object region of an object in image data;

an image processing means for erasing part of an image corresponding to said object region in said image data; and

a means for detecting an edge of a document in said image data;

wherein said object detection means detects said object region by setting an initial cotour based on said image data, and deforms said initial contour along a border of variable density in said image data so as to determine said object region, and

wherein said initial cotour is set as a straight line of fixed endpoints having a deformation tendency extending to an interior side of said document along the detected edge, said straight line being initially positioned on an exterior side of said detected edge.

2. An image processing device comprising:

an object detection means for detecting an object region of an object in image data;

an image processing means for erasing part of an image corresponding to said object region in said image data;

wherein said object detection means detects said object region by setting a specific initial contour corresponding to said image data, and deforms said initial contour along a border of variable density in said image data;

a means for detecting an edge of a document in said image data; and

a means for detecting a specific density area straddling said detected edge in said image data;

wherein said object detection means sets as an initial contour a curve through first and second intersection points of the intersection of said detected edge and a detected specific density area.

3. An image processing device in accordance with claim 2, wherein said object detection means detects said object region by detecting a boundary in said image data and setting an initial object perimeter curve based on said detected boundary.

4. An image processing device in accordance with claim 3, wherein said object detection means adjusts said initial object perimeter curve based on density differences in said image data until an adjusted object perimeter curve approximates an actual perimeter of said object region.

5. An image processing device in accordance with claim 4, wherein said image processing means erases said portion of an image corresponding to said object region in said image data by substituting a background image value for image values on an interior side of said adjusted object perimeter curve.

6. an image processing device in accordance with claim 4, wherein said initial object perimeer curve is adjusted by an iterative method until said adjusted object perimeter curve converges to an actual perimeter of said object region.

7. An image processing device comprising:

an object detection means for detecting an object region in image data which corresponds to an extraneous object in said image data, said object detection means detecting said object region by setting an initial object perimeter curve based on said image data and adjusting said initial object perimeter curve until said adjusted object perimeter curve approximates an actual perimeter of said object region;

an image processing means for erasing said object region in said image data based on said adjusted object perimeter curve; and

a means for detecting an edge of a document in said image data;

wherein said object detection means sets said initial object perimeter curve as a single straight line on an exterior side of the thus detected document edge.

8. An image processing device in accordance with claim 7, wherein said object detection means adjusts said initial object perimeter curve based on density differences in said image data until said adjusted object perimeter curve approximates an actual perimeter of said object region.

9. An image processing device in accordance with claim 7, wherein said image processing means for erasing said object region in said image data erases said object region in said image data by substituting a background image value for image values on an interior side of said adjusted object perimeter curve.

10. An image processing device in accordance with claim 7, wherein said image data corresponds to an image of a document disposed in a face upward condition.

11. An image processing device in accordance with claim 7, further comprising:

an image sensor to be disposed above a document to be read for reading image data of a document.

12. An image processing device in accordance with claim 7, wherein said initial object perimeter curve is adjusted by an iterative method until said adjusted object perimeter curve converges to an actual perimeter of said object region.

13. An image processing device comprising:

an object detection means for detecting an object region in image data which corresponds to an extraneous object in said image data, said object detection means detecting said object region by setting an initial object perimeter curve and adjusting said initial object perimeter curve until said adjusted object perimeter curve approximates an actual perimeter of said object region;

an image processing means for erasing said object region in said image data based on said adjusted object perimeter curve;

means for detecting an edge of a document in said image data; and

means for detecting a density region in said image data;

wherein said object detection means detects first and second intersection points between said detected edge of said document and said detected density region, and sets an initial object perimeter curve as a curve through said first and second intersection points.

14. An image processing device in accordance with claim 13, wherein said means for detecting a density region in said image data detects a plurality of density regions in said image data, and wherein for each of said plurality of density regions, said object detection means detects first and second intersection points between said detected edge of said document and said detected density region and sets an initial object perimeter curve as a curve through said first and second intersection points.

15. An image processing device comprising:

an object detection means for detecting an object region in image data which corresponds to an extraneous object in said image data, said object detection means detecting said object region by setting an initial object perimeter curve and adjusting said initial object perimeter curve until said adjusted object perimeter curve approximates an actual perimeter of said object region; and

an image processing means for erasing said object region in said image data based on said adjusted object perimeter curve;

wherein said initial object perimeter curve is adjusted by an iterative method until said adjusted object perimeter curve converges to an actual perimeter of said object region and wherein said initial object perimeter curve is defined by a plurality of nodes and wherein said convergence of an initial object perimeter curve is determined by minimizing a combined energy for each node, said combined energy for each node being a sum of a first energy corresponding to pixels in said close search range inside said initial object perimeter curve and a second energy corresponding to pixels in said close search range outside said initial object perimeter curve.

16. A method of processing image data comprising the steps of:

detecting an object region in image data which corresponds to an extraneous object, said detection of an object region being performed by detecting an edge of a document in said image data, setting an initial object perimeter curve as a single straight line on an exterior side of a detected document edge and adjusting said initial object perimeter curve until an adjusted object perimeter curve approximates an actual perimeter of said object region; and

erasing said object region in said image data based on said adjusted object perimeter curve.

17. A method of reading a document comprising the steps of:

detecting an object region in image data which corresponds to an extraneous object, said detection of an object region being performed by setting an initial object perimeter curve and adjusting said initial object perimeter curve until an adjusted object perimeter curve approximates an actual perimeter of said object region;

erasing said object region in said image data based on said adjusted object perimeter curve;

wherein said step of setting an initial object perimeter curve includes the steps of;

detecting an edge of a document in said image data;

detecting a density region in said image data; and

detecting first and second intersection points between said detected edge of said document and said detected density region and setting said initial object perimeter curve as a curve through said first and second intersection points.

18. A method of reading a document in accordance with claim 17, wherein said step of adjusting said initial object perimeter curve adjusts said initial object perimeter curve based on density differences in said image data until said adjusted object perimeter curve approximates an actual perimeter of said object region.

19. A method of reading a document in accordance with claim 17, wherein said step of erasing said object region in said image data erases said object region by substituting a background image value for image values on an interior side of said adjusted object perimeter curve.