GB2220743A - Measurement method and apparatus - Google Patents

Abstract

The distance between regularly spaced features such as wales and courses (Fig. 2) of knitted fabric 12 is measured by forming an image of at least a linear region containing such features, measuring the image intensity distribution (Fig. 3) along the image and fourier-analysing 15 the distribution. The method can be used for accurate repetitive measurements involving small changes, e.g. in measurement of fabric shrinkage and strain, even dynamically as part of a negative feed- back loop in fabric production or processing. The image may be formed by a TV camera 11 or a CCD array plus lenses (42 and 46, Fig 5). <IMAGE>

Description

MEASUREMENT METHOD AND APPARATUS This invention relates to the measurement of regularly occurring features in materials. Developed primarily for the rapid evaluation of the spacing of courses and wales in knitted textile fabrics, it clearly has application to a wide variety of measurements.

Traditionally, wale and course spacing in textile fabrics is estimated by using a so-called "piece glass" which is a pocket magnifying glass mounted in a folding stand of which the base comprises a square frame of side one inch or one centimetre, graduated usually in tenths. The base is placed on the textile, which is then in focus, and aligned with the courses and wales.

The numbers of courses and wales appearing in the frame are then counted.

Whilst this is a straightforward operation, it is tedious, especially if many such measurements have to be made, time-consuming and error-prone. In this latter regard, even an experienced technologist can easily be out by a whole course or wale simply as a result of not optimally positioning the frame on the material. On, say, a fourteen gauge fabric, this represents a possible error of 7%, or nearly 20% if a centimetre glass is used ("gauges" traditionally being measured in "courses per inch"). A single measurement, moreover, gives little or no information as to variability, should this be required, and multiple measurements, as noted, are tedious.

The lack of accuracy renders this measurement technique useless as a means of detecting shrinkage in fabrics, which may be smaller in magnitude than the measurement error.

The present invention provides measurement methods and apparatus which are inherently faster and easier to use than the piece glass and which provide for a higher order of accuracy.

The invention comprises a method for the measurement of regularly spaced features comprising forming an image of at least a linear region containing such features, measuring the image intensity distribution along the image and fourier-analysing the distribution.

The method can be carried out in different ways in different circumstances and using different embodiments of equipment.

In a laboratory set-up, the method can involve forming a two-dimensional video image of, say, a fabric of which the wales and/or courses are to be measured, defining a line on the video image along which the measurement is to be made (this will, of course, usually be along a course, for measurement of wales, or a wale, for measurement of courses), and measuring the image intensity distribution along the line.

Such a video image, for ease of processing, can be captured by a frame grabber and stored in a frame store, and is digitised (in available equipment in the frame grabber) for subsequent handling by a computer or microprocessor. Measurement of image intensity can be effected simply by reference to the digitised intensity level in the relevant framestore address.

The image may be enhanced by any known enhancement technique to accentuate the regularly occurring features.

Thus, for example, the image may be electronically modified by predetermining a grey-scale level and changing lighter greys to white and darker greys to black to maximise contrast. The video camera or line scan camera or CCD can have several "takes" and an average signal taken to eliminate noise.

The defined line on the two-dimensional image may be movable to select a position and orientation on the image where the measurement is to be made. Where a line scan camera or a linear CCD is used, clearly, it will be positioned to image the desired linear feature of the article being measured.

The result of the imaging along the desired line of the article, in any event, is a set of data representing image intensity, or enhanced image intensity, at points along the line, the points being pixels of the video image. This intensity distribution is then subjected to fourier-analysis, which yields the frequencies at which the signal has most energy. Its advantage lies in finding hidden periodicities in a noisy or distorted signal.

Fourier-analysis can be carried out readily and rapidly in a digital computer or'microprocessor, Where the regular feature comprises courses or wales of a knitted fabric, the principal cyclic component will clearly correspond to the spacing of such courses or wales.

Where fourier-analysis is referred to herein, the closely-related procedure of finding a Hartley transform may be substituted. "Fourier-analysis" is intended to comprehend any methodology which extracts the required periodicities from the data resulting from the imaging.

Simple mathematical manipulation using the known scale of the image and the length of the measurement line on the image will enable the frequency to be transformed into a spacing.

The method can be included as an element of a control arrangement for controlling the spacing of the said features. The measurement may be compared with a reference and an error function derived from the comparison which is applied in a negative feed-back loop to control the said spacing. The method may be included as an element of a knitting process or a process for the treatment of a knitted fabric, controlling course and/or wale spacing.

The invention also comprises apparatus for the measurement of regularly spaced features comprising imaging means forming an image of the features, image intensity measuring means measuring the image intensity distribution along an image line including the features, and fourier-analysing means automatically analyisng the measured image intensity distribution.

The imaging means may comprise a video camera or a two-dimensional charge coupled device, with means in an associated computer system for selecting a line of the image thereby produced as the measuring line. The vide camera or CCD may be associated with a frame grabber and frame store and digitiser as aforesaid.

However, a miniature arrangement, capable of substituting for the pocket piece glass, may comprise a simple line imaging device which is aligned e.g. with courses or wales. The line imaging device may comprise a linear charge coupled device, which may be incorporated in a housing containing also a source of illumination and a microprocessor into memory locations of which are put digitised signals representing the intensity distribution measured by the linear CCD, the microprocessor also containing a program fourieranalysing the distribution and outputting the result thereof in appropriate form (e.g. as a course or wale spacing in mm) to, for example, a liquid crystal display, which may be used to display also additional information such as the variance of the spacing which may be computed by the microprocessor.

Embodiments of apparatus and methods for the measurement of regularly spaced features will now be described with reference to the accompanying drawings, in which : Figure 1 is a diagrammatic representation of one embodiment of apparatus for measuring wales and courses of a knitted textile fabric; Figure 2 is a portion of a video image of the apparatus of Figure 1; Figure 3 is a portion of another video image; Figure 4 is a perspective view of another embodiment; and Figure 5 is a cross-section of the device illustrated in Figure 4.

The apparatus illustrated in Figures 1 to 3 comprises a video camera or two dimensional charge coupled device 11 aimed at a knitted textile fabric 12 of which it is desired to measure wale and course spacing. The camera or CCD 11 is connected to a frame grabber 13 and framestore 14 which are connected in turn to and under the control of a computer 15 which is controlled from a keyboard 16 and which is connected to a visual display unit (VDU) 17.

The computer 15 is programmed to accept a digitised image from the frame store 14, which is to say store the image as digitised i.e. binary data in. memory locations of the computer. The computer 15 may also be programmed then to enhance the image in known ways. For example, the image may be represented on a grey scale of 0-7, 0 representing black and 7 white. By way of enhancement, all pixels having a grey scale value of 0-3 may be reduced to 0, while all those having a value 4-7 increased to 7. The computer may also average the values of four frames to eliminate noise.

The fabric 12 may be supported on a lightbox 18 for back illumination, or may be illuminated in any other suitable fashion.

Figure 2 illustrates part of a first screen image produced by the video camera or CCD/frame grabber/ framestore/computer arrangement. The image is a magnified view of part of the fabric 12 showing the individual stitches 21 of the fabric.

A line 22 generated by programming in the computer 15 is superimposed on the image of the fabric 12. The line 22 may be moved about the image both translationally and rotationally, all in known fashion.

The line represents a measuring line and is manipulated about the screen by keyboard signals, e.g. the up and down cursor controls may shift the line up and down the screen, the east and west cursor controls covering east, west displacement and the R and L keys rotating to the right and to the left respectively.

Using these controls the line is positioned in an optimum position for measurement. The screen image of the fabric 12 will usually be of variable quality and the position with the clearest line of stitches in the course or the wale direction as appropriate can be selected by suitably positioning the line.

The length of the line 22 may correspond to a known length on the fabric 12, the scaling factor being known or determined by measurement.

One way of determining the scaling factor is to place an object of standard length on a lightbox and program the computer to evaluate the scale factor therefrom.

With the line 22 in place a further keyboard instruction to the computer 15 effects measurement of the intensity of the image along the line. In fact, the line position determines which pixels of the image are to figure in the measurement, and those pixels correspond to memory locations for the digitised values of the intensity of the image or the enhanced image.

These values are then subjected, by the computer 15, to fourier-analysis. This is a well understood operation which can be quickly carried out by a computer and yields values for the frequencies of cyclic variations in image intensity along the line 22. These values - and the fundamental frequency will be the most significant usually, so that the higher frequencies can be suppressed - can be transformed, by simple division into the length represented by the line 22, into spacings between stitches 21, which of course represent wale or course spacings whichever is appropriate.

Figure 3 shows a further image which is displayed on the screen (the stitches 21 can be left visible on the screen, the further image being superimposed, so that the correlation can be checked visually) after the measurement and calculation, where the measured intensity is plotted against the distance along the line 22, the graph being a series of peaks and troughs where each peak represents a stitch 21. The mean distance between stitches, in mm, is indicated by the side of the graph, and also, if desired, the standard deviation or variance, or any other interesting statistic.

The apparatus illustrated in Figures 1 to 3 is, of course, highly complex and sophisticated and no substitute for a piece glass, though it can give more accurate information and possibly be of considerable benefit where many evaluations of course or wale spacing have to be made accurately and rapidly.

Figures 4 and 5, however, illustrate a development of this laboratory-type equipment in which the video camera 11 is replaced by a linear CCD and everything is miniaturised so as to fit inside a pocketable unit.

These two figures illustrate a housing 41 containing a linear change coupled device 42 having, say, 64 or 128 picture elements which is connected to a microprocessor 43 with a random access memory 44 for storing a line scan of the CCD 42 and a read only memory 45 containing programming for operating the device and for performing the fourier-analysis of the line scan information. In this embodiment the image is virtual, in the sense that no visible image is made of the fabric 12 being measured, rather the intensity distribution along the CCD linescan is the virtual image.

A lens system 46 focusses the image of the fabric 12 into the CCD which is in the bottom of the housing 41 which in turn sits on a frame 47 which has a picture frame base 48 which rests on the fabric 12 during the measurement. A lamp arrangement 49 is distributed along the device running parallel to the CCD 42 and a switch 51 is depressed after the device is in place on the fabric 12 to actuate the device.

The device can, of course, be aligned with the wales or courses of the fabric 12 by suitably positioning the base 48 with long edges parallel to wales or two courses as required. The device is, of course, battery operated.

On top of the device is a liquid crystal display 52 connected to the microprocessor 43 for displaying the results of the measurement.

Claims (18)

1 A method for the measurement of regularly spaced features comprising forming an image of at least a linear region containing such features, measuring the image intensity distribution along the image and fourier-analysing the distribution.

2 A method according to claim 1, in which the image is a two-dimensional video image.

3 A method according to claim 2, in which the image is captured by a frame grabber and stored in a frame store.

4 A method according to any one of claims 1 to 3, in which the image is digitised.

5 A method according to any one of claims 1 to 4, in which the image is enhanced.

6 A method according to claim 5, in which the contrast of the image is increased.

7 A method according to claim 5 or claim 6, in which an average image is produced from multiple individual images to eliminate noise.

8 A method according to any one of claims 1 to 7., in which the fourier analysis is carried out automatically in a digital computer or microprocessor.

9 A method according to any one of claims 1 to 8, in which the regularly spaced features comprise courses or wales of a knitted fabric.

10 A method according to any one of claims 1 to 9, included as an element of a control arrangement for controlling the spacing of said features.

11 A method according to claim 10, in which the measurement is compared with a reference and an error function derived from the comparison which is applied in a negative feed-back loop to control the said spacing.

12 A method according to claim 10 or claim 11, included as an element of a knitting process or a process for the treatment of a knitted fabric, controlling course and/or wale spacing.

13 Apparatus for the measurement of regularly spaced features comprising imaging means forming an image of the features, image intensity measuring means measuring the image intensity distribution along an image line including the features, and fourier-analysing means automatically analysing the measured image intensity distribution.

14 Apparatus according to claim 13, in which the imaging means comprise a video camera or charge coupled device.

15 Apparatus according to claim 14, comprising an associated computer system including means for selecting a line of the image thereby produced as the measuring line.

16 Apparatus according to claim 13, comprising a line imaging device which can be aligned with a line including the features to be measured.

17 Apparatus according to claim 16, incorporated in a housing containing also a source of illumination and a microprocessor into memory locations of which are put digitised signals representing the intensity distribution measured by a linear charge coupled device in the housing on which is formed an image of the features to be measured, the microprocessor also containing a program fourier-analysing the distribution and outputting the result thereof.

18 Apparatus according to claim 17, comprising a liquid crystal display for the said result.