Automated apparatus for counting pillings in textile fabrics

• US 5,936,665 A
• Filed: 05/20/1997
• Issued: 08/10/1999
• Est. Priority Date: 05/22/1996
• Status: Expired due to Fees

• Current Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Original Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Inventors: Mahmood, Waqar, Dar, Iqbal M., Vachtsevanos, George J.
• Primary Examiner(s): Le, Vu
• Application Number:
  US08/859,443
• Time in Patent Office: 812 Days
• Field of Search: 348/86, 348/88, 348/91, 348/92, 348/125, 348/128, 348/138, 382/108, 382/111, 382/141, 382/144, 382/165, 382/170, 382/192, 382/194, 382/209, 382/224, 364/469.01, 364/470.01, 364/470.14, 364/470.15
• US Class Current: 348/91
• CPC Class Codes: G01N 21/8903 using a multiple detector array
• CPC Class Codes: G01N 21/8983 for testing textile webs, i.e. woven material
• CPC Class Codes: G06M 1/101 by electro-optical means
• CPC Class Codes: G06M 11/00 Counting of objects distributed at random, e.g. on a surface
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 2207/20048 Transform domain processing
• CPC Class Codes: G06T 2207/30124 Fabrics; Textile; Paper
• CPC Class Codes: G06T 7/001 using an image reference approach

Abstract

An automated system and associated method of counting pillings in textile fabrics. Images of consecutive sections of a fabric sample are captured by a CCD camera. A system processor then processes the images to enhance image quality. Pillings on the processed, captured fabric sample images are counted, and the resulting data is fuzzified to determine the membership of the data in one or more of a plurality of fabric classes. The present invention provides an objective rating system and method with repeatable accuracy for fabric samples that eliminates the inherent subjectivity associated with conventional manual visual fabric inspection methods.

Claims

1. A pilling detection system, comprising:

a pilling detection device that detects pillings which are distributed across a fabric sample;

a feed mechanism that provides the fabric sample to the pilling detection device at a predetermined feed rate; and

a processor that controls the predetermined feed rate of the feed mechanism and that assigns a class rating to the fabric sample based on comparing the number of pillings detected by the pilling detection device to stored historical classification data, said processor being programmed with fuzzy logic to determine said class rating based upon how said pillings are distributed across said fabric sample.

2. The pilling detection system of claim 1, wherein the stored historical classification data comprises a plurality of fuzzy membership values that represents historical classifications for a plurality of fabric samples of like color and texture.

3. The pilling detection system of claim 2, wherein the stored historical classification data further comprises a plurality of histograms programmed into the processor that represent a range of pilling counts associated with the plurality of sets of fuzzy membership values.

4. The pilling detection system of claim 2, further comprising a plurality of sets of fuzzy membership values each representing historical classifications for a plurality of fabric samples of like color and texture.

5. The pilling detection system of claim 4, wherein the stored historical classification data further comprises a plurality of histograms programmed into the processor that represent a range of pilling counts associated with each of the plurality of sets of fuzzy membership values.

6. The pilling detection system of claim 2, further comprising a look-up table programmed into the processor that enables the processor to determine a class rating for the test fabric sample based on a comparison of the fuzzy membership values for the number of pillings detected by the pilling detection device with predetermined quality control requirements.

7. The pilling detection system of claim 1, wherein the pilling detection device comprises a charge coupled device (CCD) array camera.

8. The pilling detection system of claim 7, further comprising a light source associated with the CCD camera that illuminates the test sample fabric for image resolution purposes.

9. The pilling detection system of claim 8, wherein the light source is an ultraviolet light source that illuminates dark-colored fabric samples.

10. The pilling detection system of claim 8, wherein the light source is a neon light source that illuminates light-colored fabric samples.

11. The pilling detection system of claim 1, further comprising an image grabber associated with the processor that records a gray scale image of the test sample fabric for pilling count purposes.

12. The pilling detection system of claim 11, wherein the processor generates a mask for each recorded gray scale image to prevent pillings from being counted more than once in consecutively recorded images.

13. The pilling detection system of claim 11, wherein the feed mechanism is operatively coupled to the image grabber and is adjustable so that a correct pilling count is obtained with minimum stepping of the feed mechanism.

14. The pilling detection system of claim 1, wherein the pilling detection device comprises a camera that has a predetermined field of view and pixel resolution.

15. An automated method for counting fabric pillings, comprising the steps of:

capturing images of consecutive sections of a fabric sample, said pillings being distributed across said fabric sample;

processing the captured images to enhance image quality;

counting pillings on the processed, captured fabric sample images; and

comparing the counted pillings to stored quality control data to provide an objective rating of the fabric sample with repeatable accuracy based upon how said pillings are distributed across said fabric sample.

16. The method of claim 15, wherein the step of comparing the counted pillings to stored quality control data comprises the step of comparing the counted pillings to stored historical pilling count data for a plurality of like fabric samples.

17. The method of claim 16, wherein the step of comparing the counted pillings to stored historical pilling count data comprises the step of:

assigning a fuzzy membership value to the pilling count for each of a plurality of class ratings.

18. The method of claim 15, wherein the step of processing the captured images comprises the steps of:

selecting a sub-image within a region of a captured image;

determining a sub-image threshold value to segment the sub-image into a background two-dimensional sub-image and a fabric two-dimensional sub-image;

transforming the fabric two-dimensional sub-image into a one-dimensional sub-image;

filtering noise from the transformed one-dimensional fabric image; and

detrending the filtered one-dimensional fabric image to account for thickness variations in the fabric sample.

19. The method of claim 18, wherein the step of counting pillings on the processed, captured fabric sample image comprises the step of:

counting the connecting segments of the detrended, filtered one-dimensional image above a pre-determined threshold.

20. The method of claim 19, further comprising the steps of:

generating a mask for the sub-image after the step of counting the connecting segments; and

comparing a mask of a current image with a mask of a previous image to ensure that same location pillings are counted only once.

21. The method of claim 15, wherein the step of comparing the number of pillings counted to stored quality control data comprises the step of implementing fuzzy logic to provide an objective rating of the sample fabric based on historical subjective ratings of like fabric samples.

22. An automated apparatus for counting pillings in textile fabrics, comprising:

a pilling detection device that detects pillings which are distributed across a fabric sample;

a feed mechanism that provides the fabric sample to the pilling detection device at a predetermined feed rate;

a processor programmed with fuzzy logic to compare the detected pilling count as distributed across said fabric sample to historical fuzzy distributions stored in the processor, assign a degree of confidence to signify the extent of membership of the pilling count to any overlapping fuzzy distributions, and assign a class rating to the sample fabric based on the extent of membership of the pilling count to overlapping fuzzy distributions and quality control parameters programmed into the processor.

23. An automated method of counting pillings in textile fabrics, comprising the steps of:

capturing images of consecutive sections of a fabric sample, said pillings being distributed across said fabric sample;

processing the captured images to enhance image quality;

counting pillings on the processed, captured fabric sample images; and

fuzzifying data generated from the step of counting pillings to determine the membership of the data in one or more of a plurality of fabric classes based upon how said pillings are distributed across said fabric sample.