WO2002057752A1

WO2002057752A1 - Method and device for examining the color and/or sheen quality of fabrics and similar materials

Info

Publication number: WO2002057752A1
Application number: PCT/EP2002/000443
Authority: WO
Inventors: Patrick Herzog
Original assignee: Color Aix Perts Gmbh
Priority date: 2001-01-21
Filing date: 2002-01-17
Publication date: 2002-07-25
Also published as: AU2002242681A1; EP1358468A1; DE10102612A1; WO2002057752A8

Abstract

The invention relates to a method and to a device for carrying out color tests on samples. The invention is characterized by using an image processing system and suitable image processing algorithms to separate and then separately measure image areas that match one another in terms of color and to compare them with reference samples, thereby obtaining reproducible measuring results that are similar to a visual color test or that at least approximate the latter to such an extent that any deviations are within the admissible tolerances for objectified test results that are obtained by visual color tests.

Description

Method and device for testing the color and / or gloss quality of fabrics and similar materials

The invention relates to a method and a device for testing the color and / or gloss quality of fabrics and similar materials

In the textile industry, in particular the clothing industry, it is customary to sample the individual items of clothing, with both their overall design and the quality of the fabrics, particularly with regard to their color appearance, being particularly important. Therefore, on the one hand, the manufacturer of the fabrics must guarantee to the processing plant for certain orders that the color quality of the fabrics is maintained; matches the color appearance of the corresponding sample parts.

The demand for compliance with a certain color quality and color consistency gains further importance if, for example, a suit jacket and corresponding trousers of different sizes have to be combined with one another (so-called "mix and match area"), or if a season has already been bought Garment to be bought another same color.

Although high-quality color measurement devices have been on the market for many years to check the color quality, the visual inspection of the photometric measurement is preferred, especially in the mix and match area.

The main reason for this is that the practice of color testing has shown that the actually objective color measurement data do not match the visual data, or do so only inadequately

BESTATIGUNGSKOPIE Correlate results. There are not small, subjective deviations here, but the deviations that occur are clearly visible and also show reproducible color differences across several observers. It can happen that a series sample is classified by the color measuring device as being absolutely within tolerance, while visually clear, intolerable color differences are determined. Conversely, it can also occur that the color measuring device detects large color distances between the respective reference area or the reference piece, whereas only a color deviation that does not exceed the tolerance specification and is therefore tolerable is determined visually.

Although visual color control by persons always has the disadvantage of the subjectivity of different observers and / or the lack of consistency of one and the same observer between different days or times of the day, with psychological aspects such as different moods of the observer as well as their reliability being added, this nevertheless presents the essential Higher expenditure of time and the subjective decision-making by the observer are the predominant alternative to color testing using known color measuring devices.

The invention is therefore initially based on the object of providing a method for carrying out color tests on templates which, while avoiding the abovementioned disadvantages, provides reproducible measurement results which are similar to those of a visual color test, but at least come so close to them that their deviations are objectified, test results that are actually visually determined lie within the permissible tolerance limits for these.

The invention is based on the knowledge that the differences that arise in the known color measurement methods or the known color measurement devices from the result of a visual assessment are based on a plurality of facts that are to be seen therein first are that the color distance formulas used here do not represent human color vision with sufficient accuracy, so that there are different tolerance thresholds which depend, for example, on the position in the color space and also on the direction of the color deviation. Since these different tolerance thresholds also differ from substance type to substance type and are also not known, they would have to be determined experimentally for each substance type. Since this is practically impossible, the difference between the achievable measurement result and the visual observation cannot be eliminated.

The differentiation between warp and weft, especially the associated gloss behavior leads to a strong dependence of a measured and visually perceived color on the angle of incidence of the illuminating light, the measurement or viewing angle and the orientation of the original, i.e. its possible rotation around an axis perpendicular to the surface of the template. This also creates non-eliminable differences between a measurement result and the visual assessment.

Color measuring devices always follow one of the standard geometries due to their structure, while a person is not able to view the templates under an exact standard geometry throughout their entire daily working hours. While color measuring devices therefore always provide an objective measurement result, the result of the visual observation is always dependent on the current viewing conditions and thus subjectively influenced or subjectively influenceable.

The gloss that is essential for the appearance of a fabric varies with the orientation as well as the lighting and measurement geometry. The exact extent of this variation must be reproduced from the series sample compared to the reference sample, otherwise the results will differ.

The light most commonly used in known color measuring devices contains only a small proportion of ultraviolet light (UV light). Daylight simulator tubes, however, usually emit a considerable amount of UV Light off. This leads to substances with optical brighteners

(fluorescent dyes), which are used in approx. one third of all yarns or

Substances are used that look different in visual observation than when measured.

Color measuring devices work either with directional or with diffuse

Light. In contrast, both the fluorescent tubes used in the visual inspection and the daylight simulators emit both directed light and a significant proportion of diffuse

Light off.

These different lighting conditions also contribute to the difference between a measurement and visual observation.

Due to the problems listed above and their influence on the measurement result, it can be seen that a reliable color measurement system for quality control of the color and / or gloss of textiles or similar materials, whose visual appearance is to be determined, must be able to, on the one hand, the To reproduce observation geometries as precisely as possible and, on the other hand, to eliminate the facts influencing the measurement result and establishing a difference between this and the visual observation with regard to their effect on the measurement result.

To solve this problem, taking into account the problems outlined above, a method is proposed which has the following method steps:

A. A white image original with a known spectral reflection factor is placed on an object table and illuminated by means of a light source whose spectral properties are known;

B. by means of a digital image recorder, preferably a multispectral image recording system with a plurality of spectral filters that can be introduced into the optical beam path, or by means of an electronically tunable filter, the white image original is recorded and digitized;

C. Each channel of the white image recording is opened when the Shutter of the image sensor made a black picture;

D. Then, instead of the white image template, the template to be measured is placed and illuminated by the light source;

E. The template is recorded and digitized by means of the digital image sensor;

F. in a data processing device, the values of the white image original and the black images are included in the digital values of the original in order to eliminate inhomogeneities and the spectral properties of the light source and to produce a corrected image;

G. the areas to be evaluated are determined using the digital image of the template;

H. Averaging of all digital values of the corrected image takes place separately for each area and channel;

I. From the averaged values of all channels, the reconstruction of the reflection spectra of the respective areas of the digitized and corrected image is carried out using a suitable method, which is based on the known spectral properties of the image recorder, the spectral filter, the white image original and the illumination;

J. the reflection spectra are multiplied separately for each area by the spectral radiation distribution of one or more types of light and by the spectral sensitivities of a human viewer (e.g. normal spectral value curves) and integrated over the visible area of the spectrum;

K. the color values XYZ obtained in this way are transformed into a suitable color space adapted to human color perception, such as CIELAB, L. The CIELAB color values are compared with the CIELAB color values of a reference template determined according to the same method and according to a color distance formula adapted to human color perception, e.g. B. DELTA-E, CIE94, CMC or CIE2000, optionally broken down according to the components contributing to the color difference, such as brightness, chroma, hue, mathematically evaluated;

M. the color difference for one or more types of light or its subcomponents is / are compared with a predetermined tolerance threshold, which may differ for the subcomponents, exceeding the tolerance threshold leading to the output of an error message.

Since not all of these procedural steps have the preceding procedural step as a prerequisite for their feasibility or effect, it is readily possible to change the order of certain procedural steps within the scope of the entire process.

Since the measuring system according to the invention is an image processing system, colored patterned substances can also be checked, in which the respective color-related image areas are separated and measured separately by suitable image processing algorithms and compared with the reference template.

Further advantages of both the method and the device used to implement it result from the following description of both the method and the device.

An apparatus for carrying out the method according to the invention has a multispectral image recording system for recording and digitizing both the white image original and the original as well as a data processing device with a memory in which comparative values of a reference template can be called up, the data processing device in turn having means for: a) the values the white picture original and the black pictures in the include digital values of the template in order to eliminate inhomogeneities and the spectral properties of the light source in order to generate a corrected image of the template; b) use the digital image of the template to determine those areas that are to be evaluated; c) averaging all digital values of the corrected image separately for each area and channel; d) to carry out the reconstruction of the reflection spectra of the respective areas of the digitized and corrected image using a suitable method based on the known spectral properties of the image recorder, the spectral filter, the white image original and the illumination, from the averaged values of all channels; e) multiply the reflection spectra separately for each area with the spectral radiation distribution of one or more types of light and with the standard spectral curves (corresponding to the spectral properties of the human eye) and integrate them over the visible area of the spectrum as well as the resulting color values XYZ in one transforming a suitable color space adapted to human color perception, such as CIELAB, f) comparing the determined CIELAB color values with the CIELAB color values of the reference template determined using the same method, and using a color spacing formula adapted to human color perception, such as e.g. B. DELTA-E, CIE94, CMC or CIE2000, if necessary broken down according to the components contributing to the color difference, such as brightness, chroma, hue, to evaluate mathematically; g) to compare the color difference for one or more types of light or its subcomponents with the previously defined tolerance threshold, which may be different for the subcomponents, and to either issue an error message if the tolerance threshold is exceeded or to show the template in true color on the screen. xxxxxxxxxxx

In a preferred embodiment, the device for carrying out the method according to the invention has a multispectral image recording system in the form of a digital image sensor of a multispectral scanner, which is provided with several light sources for illuminating an object table, which serves to record both the white image template and the template to be measured. Both the white image template and the template to be measured are preferably recorded on this in a horizontal position, but it is of course possible to change the recording position as desired. A digital camera is arranged above the object table and is assigned a device for modifying the spectral properties of the entire system. This device can be, for example, a mechanical filter changer that brings the spectral filters of the image recorder into its beam path in a predeterminable order; but it can also be an electronically tunable optical filter.

All possible technologies come into question as light sources; in addition to halogen lamps, especially fluorescent tubes. Light sources with different proportions of directed and diffuse light can also be used. Xenon lamps designed as static or flash lamps can also be used as the light source.

The spectral properties of the light source used for recording / measurement are eliminated by channel-wise white balance. For this purpose, a white image template with a known spectral reflection factor that is as homogeneous as possible, as good as possible and diffusely reflecting is recorded once or at certain intervals and used to correct the individual channels.

The position of both the light sources, the template to be measured and the camera can be varied. The respective position can be fixed for a series of measurements, or change in parts during a measurement, so that all the more information about the template to be measured, in particular its gloss behavior, is obtained.

Area sensors with, for example, 2000 x 3000 pixels can be used as digital image recorders, as can line sensors that are mechanically adjusted in order to "scan" the entire original to be measured. NEN ".

Since the template is "flat", you can use the

Multispectral scanners take color-free image acquisition.

The software can be designed in such a way that a certain area of the image is only evaluated automatically or with the help of a user after the multispectral image has been recorded and displayed on the screen, or a previously set area is recorded within the entire recording area during the recording. Since a large number of pixels and thus spectral individual measurements are available, averaging can be carried out over the respective selected area, which serves to reduce the noise, but also for averaging over a sufficiently large area of the structured template to be measured. By modifying the optics, it is also possible to adapt to roughly patterned templates.

The spectra are reconstructed using software, using one of various scientifically published methods, with only possible details having to be adapted to the specific circumstances.

With the method according to the invention, the facts listed at the outset and causative for a difference between the measurement result and the visual observation are completely or at least largely eliminated, as a result of which a sufficiently precise measurement result is achieved.

The geometry of lighting - template - measurement / camera is adjusted to the geometry of lighting - template - eye used for visual inspection.

According to the invention, a standard geometry (e.g. 45/0 or d / 89) is just as little used as any other circular or purely diffuse symmetry of the measurement.

To determine the gloss variation with the viewing or the angle of incidence of the light, the template can be rotated about a vertical or horizontal axis and then measured again, or the position of the light source can be changed between the measurements, or several light sources arranged at different locations can be switched on one after the other and used for the measurement, or the position of the image sensor can be changed between the measurements, or several image sensors can be used.

For direct comparison of the measurement with the visual control, the light used for viewing (in contrast to standard light) can be included in the mathematical color distance calculation. For better treatment of optical brighteners, either the UV light of the light source is filtered out, or the same light source is used for measurement as for the visual inspection. The light source used for the measurement contains directed and diffuse components as in the visual control; this is achieved by using the same light source as in the visual inspection or by a device that partially diffuses directed light.

Since the multispectral recording system is designed for the simultaneous processing of many (e.g. 2000 x 3000) measuring points, averaging can also take place over quite large areas, so that even patterned samples can be measured reproducibly.

The method according to the invention can be integrated "online" into the production process both in the manufacturing process of fabrics or similar materials and in the production of garments. In this way it is achieved, for example, in the manufacture of the substances that color defects can be detected even during production, so that it is not necessary to produce a certain quantity of defective substances before color defects are detected.

In the accompanying drawing, a possibility of integrating a device according to the invention into the manufacturing process of substances is shown schematically.

It shows:

1 shows a front view of the device for the continuous or area-wise checking of a fabric web. during their manufacture; 2 shows a diagrammatic representation of an integrated multispectral scanner;

In FIG. 1, a digital image sensor in the form of a multispectral scanner 3 of a multispectral image recording system is arranged downstream of the outlet point of a fabric web 1 from a production system 2. The multispectral scanner 3 has a plurality, for example 16, spectral filters, not shown, which can be introduced into its beam path in a predeterminable order.

The white image template (not shown in detail) can be placed on the fabric web 1, which also serves as an object table and is guided over a dancer roller arrangement 4. At its outlet end, the fabric web 1 is guided over a deflection roller, which is not described in more detail, and is paneled on a fabric carriage 6 by means of a rocker 5.

As can be seen from FIG. 2, the multispectral scanner 3 is assigned a plurality of light sources 7, which serve in its recording area both for illuminating the white image original and the fabric web 1. The multispectral scanner 3 is connected via a data line 8 to the computer 9 of a data processing device 10, so that the digitized and possibly transformed values of the recordings of both the white image original and the material web 1 can be fed to the computer 9 and compared with the corresponding data of a reference original. The result of the comparison and, if appropriate, also the data of relevant intermediate results can then be displayed on the screen 11 of the data processing device 10 and the corresponding data can be reproducibly stored in its memory. At the same time, the corresponding data can be printed out as a measurement protocol using a printer 12.

Since both the data of the result and the relevant interim results are stored so that they can be retrieved, they can be forwarded to the remote colorimetrically calibrated viewing stations either via online or on-demand via appropriate data transmission lines, so that they can be viewed there at the same time or later to also visually evaluate the fabric produced and, if necessary, to compare it with a reference template also displayed on a screen.

Claims

claims:

1. Procedure for testing the color and / or gloss quality of fabrics and similar materials, characterized by the following steps:

C. A black picture is taken for each channel of the white picture picture when the picture sensor is unopened;

E. The template is recorded and digitized by means of the digital image sensor;

G. the areas to be evaluated are determined using the digital image of the template; H. Averaging of all digital values of the corrected image takes place separately for each area and channel;

K. The color values XYZ obtained in this way are converted into a suitable color space adapted to human color perception, e.g. CIELAB transforms

L. The CIELAB color values are compared with the CIELAB color values of a reference template determined according to the same method and according to a color distance formula adapted to human color perception, e.g. B. DELTA-E, CIE94, CMC or CIE2000, optionally broken down according to the components contributing to the color difference, such as brightness, chroma, hue, mathematically evaluated;

M. the color difference for one or more types of light or its component parts is / are compared with a predetermined tolerance threshold, which may be different for the partial components, exceeding the tolerance threshold leading to the output of an error message.

2. The method according to claim 1, characterized in that in the case of colored patterned fabrics, color-matching image areas are separated and measured separately from one another and compared with the reference template.

3. The method according to claim 1, characterized in that the CIELAB color values of both the reference template and those of the respectively measured template are stored in the memory of the data processing device.

4. The method according to claim 1, characterized in that the white image template and the template are illuminated by means of several light sources.

5. The method according to claim 4, characterized in that the (the) angle of incidence of the light sources and / or the (the) viewing angle are changed during the measurements.

6. The method according to claim 4, characterized in that the angle of rotation is changed around the surface normal of the template during the measurements.

7. The method according to claim 1, characterized in that when comparing the measurement with the visual control, the measured spectrum of the light source used in this is included in the mathematical color distance calculation.

8. The method according to claim 1, characterized in that the measured spectrum of the daylight simulation lamp used is taken into account for measuring and for tolerance comparison of areas with optical brighteners.

9. The method according to claim 1, characterized in that the UV component of the light source used for visual comparisons and the UV component of the lamp flowing into the mathematical calculations are coordinated.

10. The method according to claim 9, characterized in that the two UV components essentially coincide.

11. The method according to claim 1, characterized in that the spectral filter of the image sensor can be introduced in a preselectable order in its beam path, or an optical filter is electronically tuned.

12. The apparatus for performing the method according to claim 1, characterized by

~ the use of a multispectral image recording system for recording and digitizing both the white image original and the original;

~ A data processing device with a memory in which comparison values of a reference template can be called up, the data processing device having means to: a) include the values of the white image template and the black images in the digital values of the template in order to generate inhomogeneities and the spectral properties of the light source for generating a eliminate the corrected image of the original; b) use the digital image of the template to determine those areas that are to be evaluated; c) averaging all digital values of the corrected image separately for each area and channel; d) to carry out the reconstruction of the reflection spectra of the respective areas of the digitized and corrected image using a suitable method based on the known spectral properties of the image recorder, the spectral filter, the white image original and the illumination, from the averaged values of all channels; e) multiply the reflection spectra separately for each area by the spectral radiation distribution of one or more types of light and by the standard spectral curves (corresponding to the spectral properties of the human eye) and integrate them over the visible area of the spectrum as well as transforming the resulting XYZ color values into a suitable color space adapted to human color perception, such as CIELAB, f) comparing the CIELAB color values determined with the CIELAB color values of the reference template determined using the same method and according to a human color perception adjusted color spacing formula such. B. DELTA-E, CIE94, CMC or CIE2000, if necessary broken down according to the components contributing to the color difference, such as brightness, chroma, hue, to evaluate mathematically; g) to compare the color difference for one or more types of light or its subcomponents with the previously defined tolerance threshold, which may differ for the subcomponents, and to either issue an error message if the tolerance threshold is exceeded, or to show the template in true color on the screen.

13. The apparatus according to claim 12, characterized in that the data of a reference digitized remotely from the device can be supplied to the memory via a data transmission line and can be displayed in color on the screen or can be stored in the memory.