Method and apparatus for analyzing fabric conditions

• US 5,125,034 A
• Filed: 12/27/1990
• Issued: 06/23/1992
• Est. Priority Date: 12/27/1990
• Status: Expired due to Fees

• Current Assignee: Tubular Textile LLC (Navis Holdings, Inc.)
• Original Assignee: Compax Corporation
• Inventors: Hudson, Rick L., Belanger, Michael G.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Stellrecht, Barry
• Application Number:
  US07/635,668
• Time in Patent Office: 544 Days
• Field of Search: 356/238, 356/242, 139/1 D, 26/51.5, 26/70, 382/1, 382/8, 364/470, 73/159, 377/3, 377/10
• US Class Current: 382/111
• CPC Class Codes: D04B 35/20 responsive to defects, e.g. holes, in knitted products
• 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

Abstract

A system for analyzing fabrics, knitted or woven, particularly to perform in-process analysis of stitch count or yarn count. A camera positioned over the fabric is exposed by means of a high speed strobe to form a freeze-frame image, which is digitized and stored for computer processing. An early processing step involves scanning the image along a large number of angularly displaced scanning axes, to determine the construction line of the fabric, which may be skewed relative to the imaging camera. The image is thereafter scanned along lines parallel to (and/or at right angles to) the construction line. A more precise and useful analysis is made possible by this technique. Color insensitivity is achieved without the use of expensive infra-red strobe devices, by using a common xenon strobe and a filter to exclude wave lengths below the near-infra-red range.

Claims

1. In the process of counting stitches per dimensional unit in a fabric, wherein a digitized freeze-frame image is made of a sample area of the fabric, and an analysis is made of the image to derive a value representative of stitch count, the improvement characterized by

(a) initially scanning said digitized image in predetermined step-wise sequences of lateral and longitudinal image increments to identify generally linear regions of said image that are substantially uniform in image pattern over an extended length and are thus identifiable as axes extending parallel to lines of construction of said fabric,

(b) analyzing the ratio of the lateral step-wise increments to the longitudinal step-wise increments to derive a value representative of the skew angle of said lines of construction with respect to the orientation of said image, and

(c) thereafter scanning successive areas of said digitized image in a pattern of lateral and longitudinal step-wise increments corresponding to said skew angle, whereby to derive a value corresponding to the number of stitches per inch of said fabric in a direction generally parallel with or generally at right angles to said lines of construction.

2. A process according to claim 1, further characterized by,

(a) said digitized image being scanned in a direction to derive a value representative of the number of stitchs per inch of said fabric in a direction known in relation to said skew angle.

3. A process according to claim 1, further characterized by,

(a) said digitized freeze-frame image is derived by an imaging camera energized by a strobe flash, and

(b) the light from said strobe flash is filtered to at least partially eliminate light in the visible wave lengths and thereby render the imaging camera less sensitive to color variations within the sample area of the fabric.

4. A process according to claim 1, further characterized by,

(a) said imaging camera being energized one or more during the obtaining each freeze-frame image,

(b) at least the first of such energizations being analyzed to determine the overall level of intensity of imaging light reaching said imaging camera,

(c) one of said light and said imaging camera being thereafter adjusted as necessary to achieve predetermined levels of image energy, and

(d) the last of such energizations being utilized to derive said digitized freeze-frame image.

5. A process according to claim 1, further characterized by,

(a) said digitized image being scanned along a plurality of successive axes, each angularly displaced from the other to ascertain which of the plurality of axes scanned most closely approximates the characteriztics of a construction line of said fabric, and

(b) thereafter scanning said digitized image with reference to the selected axis, in order to derive stitch count information along axes generally parallel with or generally at right angles to said construction line.

6. In the process of counting stitches per dimensional unit in a fabric, wherein a digitized freeze-frame image is made of a sample area of the fabric, and an analysis is made of the image to derive a value representative of stitch count, the improvement characterized by

(a) initially scanning said digitized image in predetermined stepwise sequences of lateral and longitudinal image increments along a plurality of angularly displaced axes, in order to identify the lines of construction of said fabric, and

(b) thereafter scanning successive areas of said digitized image in pattern of lateral and longitudinal step-wise increments corresponding to the orientation of said lines of construction, whereby to derive a value corresponding to the number of stitches per inch of said fabric in a direction generally parallel with or generally at right angles to said lines of construction.

7. A process according to claim 6, further characterized by

(a) prior to scanning said image, reconstructing the individual pixels of the digitized image by adjusting the gray scale value thereof with reference to the gray scale values of neighboring pixels.

8. A process according to claim 7, further characterized by

(a) the gray scale values of an individual pixel is adjusted by examining a plurality of pixels in the neighborhood of the individual pixel to determine the lowest gray scale value of the neighboring pixels, and

(b) reducing the gray scale value of the individual pixel by a value which is a function the lowest gray scale value of said neighboring pixels.