Method and system for detecting a flaw in a sample image

• US 6,295,374 B1
• Filed: 04/06/1998
• Issued: 09/25/2001
• Est. Priority Date: 04/06/1998
• Status: Expired due to Term

• Current Assignee: Omron Corporation
• Original Assignee: Integral Vision, Inc.
• Inventors: Bland, Anthony R., Robinson, David A.
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): SHERALI, ISHRAT I
• Application Number:
  US09/056,049
• Time in Patent Office: 1,268 Days
• Field of Search: 382/141-152, 382/167, 382/162, 382/209-294, 382/223, 702/35, 348/86-95, 348/125-134, 700/124, 101/181, 358/504, 358/518
• US Class Current: 382/218
• CPC Class Codes: G06T 2207/10008 from scanner, fax or copier
• CPC Class Codes: G06T 2207/30144 Printing quality
• CPC Class Codes: G06T 7/001 using an image reference approach
• CPC Class Codes: G06V 10/751 Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
• CPC Class Codes: G06V 10/993 Evaluation of the quality of the acquired pattern

Abstract

A method and system are described for use at a machine vision station to correctly identify flaws within a sample image by means of flexible template image comparison where the template image and a sample image suffer from a relative geometrical distortion. The method and system use the following iterative approach: 1) generate the template image; 2) register the sample image to the template image; 3) undertake an absolute image subtraction process; 4) identify all regions of image difference; 5) create sub-images of both the template and the sample images in the region of detected error; 6) repeat the registration step 2) using the sub-images; 7) undertake the absolute subtraction process on each of the sub-images; and 8) either identify the resulting differences as flaws or go to step 5) using even smaller sub-images. The invention has particular utility in the screen printing industry as well as other industries.

Claims

1. A method for detecting a flaw in a sample image of an object in a machine vision system including a camera and a computer, the method comprising the steps of:

a) creating a template image of the object;

b) registering the template image and the sample image;

c) subtracting the template image from the sample image to provide an error image of the sample image after performing step b);

d) analyzing the error image to identify at least one region of error of the sample image;

e) creating an individual final sample sub-image of the sample image at each identified region of error;

f) creating an individual final template sub-image of the template image at each identified region of error of the sample image;

g) registering each individual final template sub-image and each corresponding individual final sample sub-image;

h) subtracting each final template sub-image from each corresponding final sample sub-image to provide a final error image for each final sample sub-image;

i) analyzing the final error image for each final sample sub-image to identify a final sample sub-image region of error for each final sample sub-image; and

i) identifying each final sample sub-image region of error as the flaw.

2. The method as claimed in claim 1 wherein steps e) and f) are based on a preselection process.

3. The method as claimed in claim 2 wherein the preselection process is a manual preselection process.

4. The method as claimed in claim 2 wherein the preselection process is an automatic preselection process.

5. The method as claimed in claim 1 wherein steps e) and f) include the step of processing each identified region of error to create each final sub-image.

6. The method as claimed in claim 1 wherein steps e) and f) include the step of creating an initial sub-image of both the template image and the sample image in each identified region of error and wherein each final sub-image are sub-images of each initial sub-images.

7. A system for detecting a flaw in a sample image of an object at a vision station including a camera, the system comprising:

first means for creating a template image of the object;

second means for registering the template image and the sample image;

third means for subtracting the registered template image from the sample image to provide an error image of the sample image;

fourth means for analyzing the error image to identify at least one region of error of the sample image;

fifth means for creating an individual final sample sub-image and an individual template sub-image in each identified region of error, the second means registering each final sub-image of the template image and the sample image, the third means subtracting each final registered sub-image of the template image and the sample image to provide a final error image for each final sample sub-image and the fourth means analyzing each final error image of each final sample sub-image to identify a final sample sub-image region of error for each final sample sub-image; and

sixth means for identifying each final sample sub-image region of error as the flaw.

8. The system of claim 7 wherein the fifth means creates the sub-images based on a preselection process.

9. The system of claim 8 wherein the preselection process is a manual preselection process.

10. The system as claimed in claim 8 wherein the preselection process is an automatic preselection process.

11. The system as claimed in claim 7 wherein the fifth means includes means for processing each identified region of error to create each final sub-images.

12. The system as claimed in claim 7 wherein the fifth means creates an initial sub-image of both the template image and the sample image in each identified region of error and wherein each final sub-images are sub-images of each initial sub-images.

13. A method for detecting a flaw in a sample image of a geometrically distortable object in a machine vision system including a camera and a computer, the method comprising the steps of:

a) presenting a geometrically distortable object for flaw detection;

b) creating a template image of the object wherein substantially no effectively constant geometric relationship exists between the template image and the sample image;

c) registering the template image and the sample image;

d) subtracting the template image from the sample image to provide an error image of the sample image after performing step c);

e) analyzing the error image to identify a region of error of the sample image;

f) creating a final sub-image of both the template image and the sample image in the identified region of error;

g) repeating steps c) through e) using the final sub-images to identify a final region of error; and

h) identifying the identified final region of error as the flaw.

14. A system for detecting a flaw in a sample image of a geometrically distortable object at a vision station including a camera, the system comprising:

first means for creating a template image of the object wherein substantially no effectively constant geometric relationship exists between the template image and the sample image;

second means for registering the template image and the sample image;

third means for subtracting the registered template image from the sample image to provide an error image of the sample image;

fourth means for analyzing the error image to identify at least one region of error of the sample image;

fifth means for creating a final sub-image of both the template image and the sample image in the identified region of error, the second means registering the final sub-images of the template image and the sample image, the third means subtracting the final registered sub-image of the template image and the sample image to provide a final error image and the fourth means analyzing the final error image to identify a final region of error; and

sixth means for identifying the final region of error as the flaw.