Textile fabric inspection system

• US 5,774,177 A
• Filed: 09/11/1996
• Issued: 06/30/1998
• Est. Priority Date: 09/11/1996
• Status: Expired due to Fees

• Current Assignee: Milliken Research Corporation
• Original Assignee: Milliken Research Corporation
• Inventors: Lane, Jeffery Scott
• Primary Examiner(s): BRITTON, HOWARD W
• Application Number:
  US08/710,003
• Time in Patent Office: 657 Days
• Field of Search: 348/88, 348/128, 382/111, 356/429, 356/430, 356/431
• US Class Current: 348/88
• CPC Class Codes: D06H 3/08 by photo-electric or television means D06H3/12, D06H3/14, D06H3/16 take precedence
• CPC Class Codes: G01N 21/8983 for testing textile webs, i.e. woven material
• CPC Class Codes: G01N 33/367 Fabric or woven textiles optical analysis of moving sheets G01N21/86
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/20036 Morphological image processing
• CPC Class Codes: G06T 2207/30124 Fabrics; Textile; Paper
• CPC Class Codes: G06T 7/0004 Industrial image inspection
• CPC Class Codes: G06T 7/136 involving thresholding
• CPC Class Codes: H04N 7/18 Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast

Abstract

A method and apparatus for inspecting textile fabric which automatically detects defects within the field of view of a video camera. A frame of video information is captured and a digital image array is generated. A mask or operator is then used to generate a gradient image from the digital image array. The gradient image is then thresholded, dilated, optionally eroded, and then subjected to a blob analysis. A special operator is utilized in order to generate a gradient image which slow grey level transitions (e.g., oil spots) are identifiable.

Claims

1. A system for inspecting textile fabric comprising:

(a) a means for assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) a memory for storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) a means for generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions;

(d) a means for thresholding said set of numerical brightness attributes for said gradient image;

(e) a means for dilating said numerical brightness attributes that have been thresholded; and

(f) a means for blob analysis of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

2. A system for inspecting textile fabric as defined in claim 1, wherein said means for assigning numerical brightness attributes for pixels in the field of view of said video camera includes a frame grabber.

3. A system for inspecting textile fabric as defined in claim 1, wherein said memory for storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera includes a computer.

4. A system for inspecting textile fabric as defined in claim 1, wherein said means for blob analysis of said numerical brightness attributes for said pixels that have been dilated includes a means for labeling connected pixels as objects and only classifying those objects above a size threshold as defects.

5. A system for inspecting textile fabric as defined in claim 1, wherein said textile fabric is a clean room wipe.

6. A system for inspecting textile fabric as defined n claim 1, wherein said means for generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera includes a plurality of masks.

7. A system for inspecting textile fabric as defined in claim 6, wherein said plurality of masks include the following arrays:

##EQU1##

8. A system for inspecting textile fabric as defined in claim 1, wherein said means for thresholding said set of numerical brightness attributes for said gradient image includes a means for converting all numeral brightness attributes of said pixels below a threshold value to substantially a lowest value and for converting the numerical brightness attributes of said pixels above said threshold value to substantially a highest numerical brightness value.

9. A system for inspecting textile fabric as defined in claim 8, wherein said means for thresholding said set of numerical brightness attributes for said pixels in said gradient image includes a means for converting said numeral brightness attribute of said pixels that have bee converted to said substantially highest numerical brightness attribute value to said substantially lowest numerical brightness attribute value if said number of pixels that are of said substantially highest numerical brightness value are located in proximity to each other in a cluster does not exceed a size threshold.

10. A system for inspecting textile fabric as defined in claim 1, wherein said means for thresholding said set of numerical brightness attributes for said pixels in said field of view of said video camera includes a means for converting all numeral brightness attributes of said pixels below a threshold value to substantially a value of representing the color black and for converting said set of numerical brightness attributes of said pixels above said threshold value to substantially a value representing the color white.

11. A system for inspecting textile fabric as defined in claim 10, wherein said color white has a numerical brightness attribute of zero (0).

12. A system for inspecting textile fabric as defined in claim 10, wherein said color black has a numerical brightness attribute of two hundred and fifty-five (255).

13. A system for inspecting textile fabric as defined in claim 1, wherein said means for dilating said numerical brightness attributes that have been thresholded includes a mask.

14. A system for inspecting textile fabric as defined in claim 13, wherein said mask includes a three by three (3×

3) array.

15. A system for inspecting textile fabric as defined in claim 14, wherein said three by three (3×

3) array includes the following;

##EQU2##

16. A system for inspecting textile fabric as defined in claim 1, wherein said means for generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions includes a mask.

17. A system for inspecting textile fabric as defined in claim 16, wherein said mask includes a Sobel operator.

18. A system for inspecting textile fabric as defined in claim 16, wherein said mask includes a three by three (3×

3) operator array.

19. A system for inspecting textile fabric as defined in claim 16, where in said mask includes the following array:

##EQU3##

20. A system for inspecting textile fabric as defined in claim 16, wherein said mask includes the following array:

##EQU4##

21. A system for inspecting textile fabric as defined in claim 16, wherein said mask includes the following array:

##EQU5##

22. A system for inspecting textile fabric as defined in claim 16, wherein said plurality of masks include the following arrays:

##EQU6##

23. A system for inspecting textile fabric as defined in claim 16, wherein said mask includes a pivot element, a negating portion that trails said pivot element and a positive portion that is offset from said negating portion of said mask.

24. A system for inspecting textile fabric as defined in claim 23, wherein said mask includes the following array:

##EQU7##

25. A system for inspecting textile fabric comprising:

(a) a means for assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) a memory for storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) a means for generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions;

(d) a means for thresholding said set of numerical brightness attributes for said gradient image;

(e) a means for dilating said numerical brightness attributes that have been thresholded;

(f) a means for eroding said numerical brightness attributes that have been dilated; and

(g) a means for blob analysis of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

26. A system for inspecting textile fabric as defined in claim 25, wherein said means for eroding said numerical brightness attributes that have been dilated includes a mask.

27. A system for inspecting textile fabric as defined in claim 26, wherein said mask includes a three by three (3×

3) array.

28. A system for inspecting textile fabric as defined in claim 27, wherein said three by three (3×

3) array includes the following;

##EQU8##

29. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual grey level spatial transitions;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

30. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera with a frame grabber;

(b) storing as set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

31. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera with a computer;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

32. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions with a mask;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

33. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions with plurality of three by three (3×

3) operator arrays;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

34. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions with the following the three by three (3×

3) operator arrays;

##EQU9## (d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

35. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions with a mask wherein said mask includes a pivot element, a negating portion that trails said pivot element and a positive portion that is offset from said negating portion of said mask;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

36. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera for identifying gradual gray level spatial transitions with the following three by three (3×

3) operator arrays;

##EQU10## (d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

37. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image by converting all numeral brightness attributes of said pixels below a threshold value to substantially a lowest value and for converting the numerical brightness attributes of said pixels above said threshold value to substantially a highest numerical brightness value;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

38. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image by converting all numeral brightness attributes of said pixels below a threshold value to substantially a value of representing the color black and for converting the numerical brightness attributes of said pixels above said threshold value to substantially a value representing the color white;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

39. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded with a mask; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

40. A process for inspecting textile fabric comprising the steps;

of;

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded with a mask wherein said mask is a three by three (3×

3) array; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

41. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded with a mask wherein said mask is the following three by three (3×

3) array;

##EQU11## and (f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

42. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in. the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated by labeling connected pixels as objects and only classifying those objects above a size threshold as defects and thereby identifying defects in said textile fabric.

43. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded;

(f) eroding said numerical brightness attributes that have been dilated; and

(g) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defect in said textile fabric.

44. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded;

(f) eroding said numerical brightness attributes that have been dilated with a mask; and

(g) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

45. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded;

(f) eroding said numerical brightness attributes that have been dilated with a three by three (3×

3) array; and

(g) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

46. A process for inspecting textile fabric comprising the steps:

of;

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said stet of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded;

(f) eroding said numerical brightness attributes that have been dilated with the following three by three (3×

3) array;

##EQU12## and (g) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

47. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image by converting said numeral brightness attributes of said pixels that have been converted to said substantially highest numerical. brightness attribute value to said substantially lowest numerical brightness attribute value if said number of pixels that are of said substantially highest numerical brightness value are located in proximity to each other in a cluster does not exceed a size threshold;

(e) dilating said numerical brightness attributes for said pixels that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said textile fabric.

48. A process for inspecting clean room wipes comprising the step of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated and thereby identify defects in said clean room wipes.

49. A process for inspecting textile fabric comprising the steps of:

(a) assigning numerical brightness attributes for pixels in the field of view of a video camera;

(b) storing a set of said numerical brightness attributes for said pixels in said field of view of said video camera;

(c) generating a gradient image from said set of numerical brightness attributes for said pixels in said view of said video camera;

(d) thresholding said set of numerical brightness attributes for said gradient image;

(e) dilating said numerical brightness attributes that have been thresholded; and

(f) analyzing blobs of said numerical brightness attributes for said pixels that have been dilated by converting said numeral brightness attributes of said pixels that have been converted to said substantially highest numerical brightness attribute value to said substantially lowest numerical brightness attribute value if said number of pixels that are of said substantially highest numerical brightness value are located in proximity to each other in a cluster does not exceed a size threshold and thereby identify defects in said textile fabric.