Method for determining the profile of a bound document with structured light

• US 5,835,241 A
• Filed: 05/30/1996
• Issued: 11/10/1998
• Est. Priority Date: 05/30/1996
• Status: Expired due to Term

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Saund, Eric
• Primary Examiner(s): Moore, David K.
• Assistant Examiner(s): BRINICH, STEPHEN M
• Application Number:
  US08/655,520
• Time in Patent Office: 894 Days
• Field of Search: 358/493, 358/488, 358/296, 358/474, 358/466, 358/445-446, 358/443, 355/25, 382/318-319, 382/203-204
• US Class Current: 358/488
• CPC Class Codes: H04N 1/047 Detection, control or error compensation of scanning velocity or position H04N1/0402 and H04N1/17 take precedence
• CPC Class Codes: H04N 1/195 the array comprising a two-dimensional array or a combination of two-dimensional arrays
• CPC Class Codes: H04N 1/387 Composing, repositioning or otherwise geometrically modifying originals
• CPC Class Codes: H04N 2201/02425 Self-adjusting arrangements, e.g. compensating for temperature fluctuations
• CPC Class Codes: H04N 2201/02439 Positioning method
• CPC Class Codes: H04N 2201/03162 Original guide plate
• CPC Class Codes: H04N 2201/0434 specially adapted for scanning pages of a book
• CPC Class Codes: H04N 2201/04703 using the scanning elements as detectors, e.g. by performing a prescan
• CPC Class Codes: H04N 2201/04717 on the scanned sheet, e.g. a reference sheet
• CPC Class Codes: H04N 2201/04729 in the main-scan direction
• CPC Class Codes: H04N 2201/04731 in the sub-scan direction
• CPC Class Codes: H04N 2201/04734 Detecting at frequent intervals, e.g. once per line for sub-scan control
• CPC Class Codes: H04N 2201/04787 by changing or controlling the addresses or values of pixels, e.g. in an array, in a memory, by interpolation
• CPC Class Codes: H04N 2201/04789 in the main-scan direction
• CPC Class Codes: H04N 2201/04791 in the sub-scan direction
• CPC Class Codes: H04N 2201/04794 Varying the control or compensation during the scan, e.g. using continuous feedback or from line to line

Abstract

An overhead scanning system records pages from bound documents in an upright and open condition. A light stripe is projected onto each page of the bound document to measure the non-planar deformations of each page. The scanning system records a first image with the light stripe and a second image without the light stripe. After identifying a set of pixels representing the light stripe in the first image, the set of pixels is smoothed to minimize large transitions between scan lines. Subsequently, a crease is identified in the set of pixels that represents the spine of the bound document. The pixels forming the crease are replaced with pixels that are defined using a parametric model that approximates the spine region of a bound document. This substitution insures that the spine region of a bound document is accurately depicted before determining a page shape transform with the identified set of pixels representing defining the projected light stripe. An image correction system de-warps the second image using the page shape transform and a perspective transform to correct distortion due to the pages of the bound document being curved relative to the support plane of the system. The de-warped image is determined by "polling" location in the second image for output values.

Claims

1. In a scanning apparatus having an image acquisition system for recording images and a light stripe projector for projecting a light stripe, a method for determining a shape of the bound document positioned in an imaging area of the scanning apparatus, comprising the steps of:

recording, with the image acquisition system, an image of the bound document having a light stripe projected thereacross by the light stripe projector, the image being defined by an array of scan lines, with each scan line having pixels with an intensity value;

adaptively thresholding the pixels forming the image to identify pixels defining part of the light stripe, said adaptive thresholding step compensating for non-uniform illumination across the bound document recorded by the image acquisition system;

identifying a pixel starting location that forms part of the light stripe on a first scan line near a midpoint of the image;

tracking, from the pixel starting location, a first set of pixels that further define the light stripe in the image, the first set of pixels representing a trace of the light stripe projected across the bound document;

smoothing the first set of pixels to minimize large transitions between scan lines;

detecting a crease location along the first set of pixels that identifies a spine region of the bound document; and

responsive to said detecting step, modifying the first set of pixels representing the trace, said modifying step replacing pixels near the crease location with a second set of pixels that are determined using a parametric model that approximates the spine region of the bound document.

2. The method according to claim 1, wherein said step of adaptively thresholding the image includes the steps of:

identifying pixels in the image with intensity values that vary in intensity from neighboring pixels by a threshold value, and

setting the intensity values of those pixels in the image identified by said adaptive thresholding step to zero.

3. The method according to claim 1, wherein said step of defining a starting location includes the step of identifying a neighborhood of pixels having a non-zero intensity value.

4. The method according to claim 3, wherein said step of identifying a starting pixel location includes the steps of:

defining a neighborhood of pixels in the image; and

computing a centroid of the neighborhood of pixels.

5. The method according to claim 4, further comprising the step of centering the neighborhood of pixels at a pixel location defined by the centroid.

6. The method according to claim 5, further comprising the step of repeating said computing step and said centering step until the neighborhood of pixels does not change.

7. The method according to claim 6, wherein said centroid is an average value of the intensity values in the neighborhood of pixels.

8. The method according to claim 1, wherein said tracking step tracks in a first direction and a second direction one scan line at a time.

9. The method according to claim 1, wherein said tracking step further comprises the steps of:

identifying, within a predefined area in the image, pixels a small distance from the pixel starting location on a second scan line having a non-zero intensity value; and

evaluating a centroid of the identified pixels with a non-zero intensity value on the second scan line.

10. The method according to claim 9, further comprising the steps of:

growing the predefined area;

identifying, within the grown predefined area, pixels a small distance from the pixel starting location on the second scan line having a non-zero intensity value; and

evaluating a centroid of the identified pixels with a non-zero intensity value on the second scan line.

11. The method according to claim 10, further comprising the step of abandoning said growing step at a predetermined maximum area.

12. The method according to claim 1, further comprising the step of transforming the set of pixels from a pixel coordinate system to a height coordinate system.

13. The method according to claim 1, wherein said smoothing step smoothes the set of pixels with a Gaussian smoothing kernel.

14. The method according to claim 1, wherein said detecting step includes the step of assessing whether the trace contains a pixel that identifies the spine region of the bound document.

15. The method according to claim 14, wherein said assessing step includes the steps of:

measuring in the image changes in height between each scan line along the trace of the light stripe projected across the bound document; and

selecting a pixel identifying a maximum measured change in height, the selected pixel identifying the spine region of the bound document.

16. The method according to claim 1, wherein said modifying step includes the step of fitting a parabola to the pixels defining the trace of the light stripe in the spine region of the bound document.

17. The method according to claim 1, wherein said detecting step identifies the crease pixel at a scan line in the image.

18. The method according to claim 1, wherein said tracking step rejects pixels that are exist apart from pixels in the first set of pixels.

19. A method for scanning a sequence of non-planar original images of a bound document, comprising the steps of:

recording geometry data and raw image data for each image in the sequence of non-planar original images, said recording step including the steps of;

projecting a light stripe on a non-planar original image,

acquiring geometry data with an image acquisition system, and

acquiring raw image data with the image acquisition system; and

correcting the recorded raw image data for each image in the sequence of non-planar original images with the recorded geometry data, said correcting step performing, for each image in the sequence of non-planar original images, the steps of;

identifying a pixel starting location that forms part of the light stripe on a scan line near a midpoint of the geometry data,

tracking, from the pixel starting location, a set of pixels that further define the light stripe in the image,

smoothing the set of pixels to minimize large transitions between scan lines,

converting the set of pixels, which represent a trace of the light stripe projected across the bound document, to a page shape transform, and

de-warping the recorded image data with the page shape transform to provide corrected image data.

20. The method according to claim 19, wherein said recording step is completed before initiating said correcting step.

21. The method according to claim 19, further comprising the step of storing the geometry data and the raw image data in a memory.

22. The method according to claim 21, wherein said correcting step is delayed until the raw image data is displayed on a terminal.

23. The method according to claim 21, wherein said correcting step is delayed until the raw image data is reproduced as hard copy.

24. The method according to claim 19, further comprising the step of recording calibration data with the image acquisition system.

25. The method according to claim 24, wherein said correcting step includes the step of determining a perspective transform with the calibration data, the perspective transform defining a relationship, between a two-dimensional coordinate system of the image acquisition system and a three-dimensional coordinate system of the bound document.

26. The method according to claim 25, wherein said de-warping step includes the step of transforming the image data with the perspective transform.

27. The method according to claim 19, wherein said recording step and said correcting step are performed in parallel for each image in the sequence of non-planar original images.

28. The method according to claim 19, wherein said recording step and said correcting step are performed sequentially for each image in the sequence of non-planar original images.