Imaging-based methods for detecting and measuring defects in extruded cellular ceramic articles

• US 9,996,766 B2
• Filed: 04/27/2016
• Issued: 06/12/2018
• Est. Priority Date: 05/01/2015
• Status: Active Grant

• Current Assignee: Corning Incorporated
• Original Assignee: Corning Incorporated
• Inventors: Richard, DiRisio Carlo, Charmaine, Frey Julie, Sluder, Gehl Kristine, Madara, Russell Wayne, Sim, Hak Chuah, Zhang, Yeqing
• Primary Examiner(s): Chawan, Sheela C
• Application Number:
  US15/139,931
• Publication Number:
  US 20160321797A1
• Time in Patent Office: 776 Days
• Field of Search: 382100, 382141, 382152, 382144, 382118, 382147, 382274, 382275, 382278, 382145, 382172, 382191, 382192, 382193, 382194, 348 92, 348125, 348128, 348133, 348134, 356601, 356337, 356437, 356402, 356608, 356602, 3562371, 3562372, 3562373, 3562374, 3562375, 3562376, 428116, 428131, 428156, 428174, 428188, 428593
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06T 2207/30164 Workpiece; Machine component
• CPC Class Codes: G06T 7/0006 using a design-rule based approach

Abstract

Imaging-based methods for detecting defects in an extruded cellular ceramic article having a web array are disclosed. The methods may include capturing a digital image of the web array to establish an intensity-based initial web array representation. The methods also may include performing an intensity threshold process on the initial web array representation to define an intensity-based high-resolution web array representation having sub-pixel resolution, and performing piecewise cubic spline fits to define corresponding web skeletons. The method may also include comparing intensities of the high-resolution web array representation along the web skeletons to a threshold intensity to determine the defect locations and sizes. The methods can be applied to determining the size and location of skin defects as well.

Claims

1. A method of inspecting an extruded cellular ceramic article for defects, the extruded cellular ceramic article having a web array comprising webs that define an array of cells, the method comprising:

establishing, from a digital image that comprises pixels each having an intensity value, an intensity-based initial web array representation of the web array that does not have sub-pixel resolution;

performing an intensity threshold process on the initial web array representation to define an intensity-based high-resolution web array representation of the web array that has sub-pixel resolution;

defining corresponding web skeletons for each of the webs by applying piecewise cubic spline fits between knot locations; and

comparing intensities of the high-resolution web array representation along the web skeletons to a threshold intensity to determine the location and the size of any defects that may be present in or on the article.

2. The method according to claim 1, wherein the establishing comprises defining cell contours by performing a threshold intensity process on the digital image and then establishing from the cell contours the intensity-based initial web array representation.

3. The method according to claim 1, wherein each web has a thickness, and wherein in the comparing, the intensity value at a given location of one of the web skeletons is a maximum intensity of intensity values over the width of the web at the given location.

4. The method according to claim 1, wherein the defining occurs after the performing, and the method further comprises after the performing but prior to the defining, performing a filtering operation that reduces or eliminates debris from the high-resolution web array representation.

5. The method according to claim 1, wherein the knot locations are determined by dividing a given web into segments that are between three and seven pixels long and that comprise edge pair midpoints, and averaging the edge pair midpoints within each segment into the knot location for the segment.

6. The method according to claim 1, further comprising determining center post locations in the web array, and performing the piece-wise cubic spline fits between adjacent center post locations.

7. The method according to claim 1, further comprising determining for each web a web curvature, and comparing the determined web curvature to a maximum acceptable web curvature.

8. The method according to claim 1, further comprising:

calculating a contrast value for each point along each of the web skeletons; and

comparing the contrast to a threshold contrast value in determining the location of the at least one defect.

9. The method according to claim 1, comprising defining the threshold intensity as a weighted sum of a local cell foreground intensity threshold T_Cell and a global web array foreground intensity threshold T_NK.

10. The method according to claim 9, wherein the threshold intensity is given as T_WG and wherein the weighted sum is defined by T_WG=0.8T_Cell+0.2T_NK.