Image processing apparatus, image processing method, image processing program, printed matter inspection apparatus, printed matter inspection method and printed matter inspection program

• US 7,684,625 B2
• Filed: 11/12/2003
• Issued: 03/23/2010
• Est. Priority Date: 12/20/2002
• Status: Expired due to Fees

• Current Assignee: Fuji Xerox Company Limited (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Xerox Company Limited (Fujifilm Holdings Corporation)
• Inventors: Hyoki, Kenji
• Primary Examiner(s): Mehta; Bhavesh M
• Assistant Examiner(s): Thomas; Mia M
• Application Number:
  US10/704,570
• Publication Number:
  US 20040146202A1
• Time in Patent Office: 2,323 Days
• Field of Search: 382/112, 382/209, 382/217, 358/1.2, 358/3.26
• US Class Current: 382/217
• CPC Class Codes: G06T 2207/30144 Printing quality
• CPC Class Codes: G06T 3/02 Affine transformations for image registration G06T3/147; for image mosaicing G06T3/4038
• CPC Class Codes: G06T 5/80 Geometric correction
• CPC Class Codes: G06T 7/001 using an image reference approach
• CPC Class Codes: G06T 7/30 Determination of transform parameters for the alignment of images, i.e. image registration

Abstract

A split image generation portion generates split images. A split projected waveform generation portion generates split projected waveforms from the split images respectively. A projected waveform synthesizer portion generates projected waveforms by combining the split projected waveforms. A feature volume extraction portion extracts feature volumes from the generated projected waveforms. A feature volume correlation portion correlates feature volume data of a target image with feature volume data of a reference image. A correction value calculation portion calculates correction values in accordance with geometrical distortion between the reference image and the target image on the basis of a result of the correlation and generates an n-dimensional lookup table. An image collation portion collates the target image with the reference image in consideration of the geometrical distortion by using the lookup table.

Claims

1. An image processing apparatus for collating a target image as a subject of inspection with a reference image, comprising:

a projected waveform generation unit that generates a projected waveform of the reference image and a projected waveform of the target image;

a feature volume extraction unit that extracts feature volumes from the projected waveform of the reference image and the projected waveform of the target image and generates feature volume data of the reference image and feature volume data of the target image, respectively;

a feature volume correlation unit that correlates the feature volume data of the target image with the feature volume data of the reference image;

a correction value calculation unit that calculates correction values in accordance with geometrical distortion between the reference image and the target image based on a result of correlation of the feature volume data of the target image with the feature volume data of the reference image; and

a collation unit that collates the target image with the reference image by using the correction values.

2. An image processing apparatus according to claim 1, wherein:

the correction value calculation unit generates an n-dimensional lookup table in which the correction values are stored as data and which is used for correcting the geometrical distortion; and

the collation unit uses the n-dimensional lookup table for performing collation.

3. An image processing apparatus according to claim 1, wherein:

the correction value calculation unit generates an n-dimensional lookup table in which the correction values and data interpolated between the correction values are stored and which is used for correcting the geometrical distortion; and

the collation unit uses the n-dimensional lookup table for performing collation.

4. An image processing apparatus according to claim 1, wherein

the feature volume correlation unit correlates the feature volume data of the target image with the feature volume data of the reference image using DP matching.

5. An image processing apparatus according to claim 1, wherein

the feature volume correlation unit correlates respective elements of the feature volume data of the target image with the feature volume data of the reference image successively.

6. An image processing apparatus according to claim 1, wherein

the projected waveform generation unit includes:

a split image generation unit that generates split images by splitting an image into m×

n blocks;

a split projected waveform generation unit that generates projected waveforms from the split images; and

a projected waveform synthesizer unit that generates a projected waveform of the image or a partial image larger in area than one of the split images by combining the projected waveforms generated from one or more of the split images respectively.

7. An image processing apparatus comprising:

a split image generation unit that generates split images by splitting an image into m×

n blocks;

a split projected waveform generation unit that generates projected waveforms from the split images; and

a projected waveform synthesizer unit that generates a projected waveform of the image or a partial image larger in area than one of the split images by combining the projected waveforms generated from one or more of the split images respectively; and

a feature volume extraction unit that extracts feature volumes from the projected waveform of the image and generating feature volume data of the image.

8. An image processing apparatus according to claim 7, comprising:

an n-dimensional lookup table containing correction values for correcting geometrical distortion in accordance with coordinates; and

a collation unit that collates a target image with a reference image by using the correction values stored in the n-dimensional lookup table.

9. An image processing method for collating a target image as a subject of inspection with a reference image, wherein a processor performs the method comprising:

generating a projected waveform of the reference image and a projected waveform of the target image;

extracting feature volumes from the projected waveform of the reference image and the projected waveform of the target image and generating feature volume data of the reference image and feature volume data of the target image, respectively;

correlating the feature volume data of the target image with the feature volume data of the reference image;

calculating correction values in accordance with geometrical distortion between the reference image and the target image based on a result of correlation of the feature volume data of the target image with the feature volume data of the reference image; and

collating the target image with the reference image by using the correction values.

10. An image processing method according to claim 9, further comprising:

generating an n-dimensional lookup table in which the correction values are stored as data and which is used for correcting the geometrical distortion; and

using the n-dimensional lookup table to collate the target image with the reference image.

11. An image processing method according to claim 9, further comprising:

generating an n-dimensional lookup table on which the correction values and data interpolated between the correction values are held and which is used for correcting the geometrical distortion; and

using the n-dimensional lookup table to collate the target image with the reference image.

12. An image processing method according to claim 9, wherein

DP matching is used for correlating the feature volume data of the target image with the feature volume data of the reference image.

13. An image processing method according to claim 9, wherein

elements of the feature volume data of the target image and the feature volume data of the reference image are collated successively to thereby perform correlation of the feature volume data of the target image with the feature volume data of the reference image.

14. An image processing method according to claim 9, wherein

the generating the projected waveform of the reference image and the projected waveform of the target image comprises:

generating split images by splitting an image into m×

n blocks;

generating projected waveforms from the split images; and

generating a projected waveform of the image or a partial image larger in area than one of the split images by combining the projected waveforms generated from one or more of the split images respectively.

15. An image processing method wherein a processor performs the method comprising:

generating split images by splitting an image into m×

n blocks;

generating projected waveforms from the split images; and

generating a projected waveform of the image or a partial image larger in area than one of the split images by combining the projected waveforms generated from one or more of the split images respectively; and

extracting feature volumes from the projected waveform of the image and generating feature volume data of the image.

16. An image processing method according to claim 15, comprising:

generating an n-dimensional lookup table containing correction values for correcting geometrical distortion in accordance with coordinates; and

collating a target image with a reference image by using the correction values contained in the n-dimensional lookup table.

17. A computer-readable medium storing a computer program for collating a target image as a subject of inspection with a reference image, the computer program comprising:

generating a projected waveform of the reference image and a projected waveform of the target image;

extracting feature volumes from the projected waveform of the reference image and the projected waveform of the target image and generating feature volume data of the reference image and feature volume data of the target image, respectively;

correlating the feature volume data of the target image with the feature volume data of the reference image;

calculating correction values in accordance with geometrical distortion between the reference image and the target image based on a result of correlation of the feature volume data of the target image with the feature volume data of the reference image; and

collating the target image with the reference image by using the correction values.

18. A computer-readable medium storing a computer program according to claim 17, wherein:

the calculating correction values generates an n-dimensional lookup table in which the correction values are stored as data and which is used for correcting the geometrical distortion; and

the collating the target image with the reference image uses the n-dimensional lookup table for performing collation.

19. A computer-readable medium storing a computer program according to claim 17, wherein:

the calculating the correction values generates an n-dimensional lookup table in which the correction values and data interpolated between the correction values are stored and which is used for correcting the geometrical distortion; and

the collating the target image with the reference image uses the n-dimensional lookup table for performing collation.

20. A computer-readable medium storing a computer program according to claim 17, wherein

the correlating the feature volume data correlates the feature volume data of the target image with the feature volume data of the reference image using DP matching.

21. A computer-readable medium storing a computer program according to claim 17, wherein

the correlating the feature volume data of the target image with the feature volume data of the reference image correlates respective elements of the feature volume data of the target image with the feature volume data of the reference image successively.

22. A computer-readable medium storing a computer program according to claim 17, wherein

the generating the projected waveform of the reference image and the projected waveform of the target image comprises:

generating split images by splitting an image into m×

n blocks;

generating projected waveforms from the split images; and

generating a projected waveform of the image or a partial image larger in area than one of the split images by combining the projected waveforms generated from one or more of the split images respectively, so as to generate the projected waveform of the reference image and the projected waveform of the target image respectively.

23. A computer-readable medium storing a computer program, the computer program comprising:

generating split images by splitting an image into m×

n blocks;

generating projected waveforms from the split images; and

generating a projected waveform of the image or a partial image larger in area than one of the split images by combining the projected waveforms generated from one or more of the split images respectively; and

extracting feature volumes from the projected waveform of the image and generating feature volume data of the image.

24. A computer-readable medium storing a computer program according to claim 23, the computer program comprising:

generating an n-dimensional lookup table containing correction values for correcting geometrical distortion in accordance with coordinates; and

collating a target image with a reference image by using the correction values contained in the n-dimensional lookup table.

25. A printed matter inspection apparatus for inspecting printed matter, comprising:

a reference image storage unit that stores a reference image used for forming the printed matter;

an image reader unit that reads an image of the printed matter as a target image which is a subject of inspection;

a projected waveform generation unit that generates a projected waveform of the reference image and a projected waveform of the target image;

a feature volume extraction unit that extracts feature volumes from the projected waveform of the reference image and the projected waveform of the target image and generates feature volume data of the reference image and feature volume data of the target image, respectively;

a feature volume correlation unit that correlates the feature volume data of the target image with the feature volume data of the reference image;

a correction value calculation unit that calculates correction values in accordance with geometrical distortion between the reference image and the target image based on a result of correlation of the feature volume data of the target image with the feature volume data of the reference image; and

a collation unit that collates the target image with the reference image by using the correction values.

26. A printed matter inspection method for inspecting printed matter, wherein a processor performs the method comprising:

reading an image of the printed matter as a target image which is a subject of inspection;

generating a projected waveform of a reference image and a projected waveform of the target image;

extracting feature volumes from the projected waveform of the reference image and the projected waveform of the target image and generating feature volume data of the reference image and feature volume data of the target image, respectively;

correlating the feature volume data of the target image with the feature volume data of the reference image;

calculating correction values in accordance with geometrical distortion between the reference image and the target image based on a result of correlation of the feature volume data of the target image with the feature volume data of the reference image; and

collating the target image with the reference image by using the correction values.

27. A computer-readable medium storing a computer program for inspecting printed matter, the computer program comprising:

reading an image of the printed matter as a target image which is a subject of inspection;

storing an image used for forming the printed matter as a reference image;

generating a projected waveform of the reference image and a projected waveform of the target image;

extracting feature volumes from the projected waveform of the reference image and the projected waveform of the target image and generating feature volume data of the reference image and feature volume data of the target image, respectively;

correlating the feature volume data of the target image with the feature volume data of the reference image;

calculating correction values in accordance with geometrical distortion between the reference image and the target image based on a result of correlation of the feature volume data of the target image with the feature volume data of the reference image; and

collating the target image with the reference image by using the correction values.