WO2002086430A2

WO2002086430A2 - Color comparison method and apparatus

Info

Publication number: WO2002086430A2
Application number: PCT/GB2002/001888
Authority: WO
Inventors: Richard May
Original assignee: Color Quest Limited
Priority date: 2001-04-24
Filing date: 2002-04-24
Publication date: 2002-10-31
Also published as: WO2002086430A3; AU2002307901A1

Abstract

A method of colour comparison comprising the steps of scanning (A) sample to Establish components of light broken down into a predetermined number of points (each point relating to the colour of light reflected from the sample) across a colour spectrum; storing the value of each point as a set of first values ('Spectral Data'); deriving from the Spectral Data a second set of values ('LAB Data'); comparing the LAB Data of a particular point with a selected standard; establishing coincidence, or of any difference, between the LAB Data and the selected standard; establishing a working value for the sample based on coincidence 02 on any differences up to a predetermined value established during the comparison step for the majority, if not all, of the predetermined number of points; making use of an algorithm depending on the sample comparing the working value with comparable colour data identify a particular product having a predetermined working colour relationship with the sample.

Description

COMPARISON METHOD AND APPARATUS

TECHNICAL FIELD

This invention relates to a comparison method and apparatus. It is particularly, but not exclusively, concerned with scanning a source material, such as a furnishing material such as a carpet in order to establish a requirement, typically for another furnishing material such as one selected from a group comprising paint, flooring, fabric, wall covering, ceramic material or other material, having a pre-determined relationship with the source material. The invention is particularly concerned with colour comparison between material of different types.

BACKGROUND ART

In our first co-pending application UK 2346964 there is described a method of generating colour based information originating from a sample characterised by the^" steps of: putting the sample into a condition where it can be scanned to provide consistent colour and light information in the form of parameters (such as information relating to hue, tone and light intensity); scanning the sample to establish first value/s of the required parameter /s according to a given first range of such parameters; generating from the first value/s second value/s of the parameter /s lying in a second range of such parameters either on the basis of a one-to-one relationship between at least a part of the first range and the second range or on the basis of some other relationship or relationships between the first and second range; displaying some or all of the first value/s or the second value/s or of functions of the first or second values or combinations of the first and second values or of functions thereof either directly or in the form of a virtual or other image. There is further described apparatus for providing information relating to a sample characterised by: a locating means adapted to locate a sample from which information is required; a scanner mounted relative to the locating means and adapted so as to be able to scan a sample located by the locating means and to provide as output information relating to the colour of the located sample; a processor adapted to receive output information from the scanner in relation to a first range of such parameters; to generate second value/s of the parameter /s in terms of a second range of such parameters either on the basis of a one-to-one relationship between the first range and the second range or on the basis of some other relationship or relationships between the first and second range; and display or transmission means whereby some or all of the first values or the second values or functions of the first or second values or combinations of the first and second values or of functions thereof can be displayed or transmitted either as information or in the form of a virtual or other image.

In our second co-pending application UK 2,349,206 there is described a method of providing information relating to the appearance of a first coloured article falling in the family of the type described to enable a second coloured article from the family to be selected or identified which can be related in a predetermined way to the first characterised by the steps of: scanning the first article to establish colour and light information and thereafter to provide an output in the form of derived information giving at least one of the parameters defining hue, tone and light intensity; relating the derived information to established parameter or parameters located in one or more ranges of such parameters to identify the required parameter or parameters of the second article is either by direct correspondence or by way of a predetermined range of such parameters such as provided by a colour circle or a range of colours in the visible light spectrum; the relationship between the colour/light information of the second article relative to the first being one being either identity, partial identity or a relationship such as by being a complementary or a contrasting colour. The invention further comprises apparatus for undertaking the method characterised by: the step of locating means adapted to locate the first article; a scanner mounted relative to the locating means and adapted to scan a first article and to provide the derived output information; a processing means adapted to receive the derived information and to relate the established parameter or parameters to one or more ranges of such parameters either by direct comparison with a physical sample or with entries in a range of such parameters held as a data base to identify the required parameter or parameters of the second article as aforesaid. DISCLOSURE OF INVENΉON

According to a first aspect of the present invention there is provided a method of colour comparison relating to a sample of a furnishing material comprising the steps of:

1 scanning the sample to establish components of light broken down into a predetermined number of points (each point relating to the colour of light reflected from the sample) across a colour spectrum extending over the light range from infra-red to ultra-violet);

2 storing the value of each point obtained from the scanning step as a set of first values ('Spectral Data') in a first data store; Spectral Data effectively giving a two dimensional view of the colour at each point;

3 deriving from the Spectral Data a second set of values ('LAB Data') in a second data store; LAB Data effectively giving a position of the colour at a point within a three dimensional colour space;

4 comparing the LAB Data of a particular point with a selected standard; establishing coincidence, or of any difference, between the LAB Data and the selected standard;

5 establishing a working value for the sample based on coincidence or any differences up to a predetermined value established during the comparison step for the majority, if not all, of the predetermined number of points making use of an algorithm depending on the sample of furnishing material involved (that is to say a given algorithm for carpets; a further algorithm for paint, yet another algorithm for wall paper, and so on);

6 comparing the working value with comparable colour data for another furnishing material product of a different type to that of the sample in order to identify a particular product having a predetermined working colour relationship with the sample.

According to a first preferred version of the first aspect of the present invention the predetermined number of points in the scanning step is 16. According to a second preferred version of the present invention or of the first preferred version thereof the product is selected from a group comprising carpet, paint, flooring, fabric, wall covering, ceramic material.

According to a third preferred version of the present invention or of any preceding preferred version thereof the step of comparing the working value with comparable colour data for a product is directed to a straight colour match between the sample and the product.

According to a fourth preferred version of the first aspect of the present invention or of the first or second preferred versions thereof the step of comparing the working value with comparable colour data for a product is directed to a relationship other than a straight colour match (such as a complementary colour relationship) between the sample and the product.

According to a fifth preferred version of the present invention or of any preceding preferred version thereof including the provision of one or more steps wherein comparison can be made between existing colour scales (such as Chromatone (RTM), Dulux (RTM), Pantone (RTM), NCS) or generated colour scales.

According to a sixth preferred version of the first aspect of the present invention or of any preceding preferred version thereof there is provided a further step, following the step of comparing the working value with comparable colour data for a product in order to identify a particular product, of establishing from an existing database information relating to the particular product such as source, location, specification or price.

According to a seventh preferred version of the first aspect of the present invention or of any preceding preferred version thereof wherein in the step establishing a working value for the sample based coincidence or of any differences established during the comparison step for the majority, if not all, of the predetermined number of points there is generated a listing showing working values giving zero difference or differences within a predetermined range or band of ranges in the case of a given sample.

According to a second aspect of the present invention there is provided apparatus for undertaking a method according to the first aspect or of any preferred version thereof.

The purpose of the software used is to drive a piece of hardware called a spectro- photometer and process the information. The spectrophotometer is an optical measurement device which uses a very powerful light to illuminate the sample and receive reflected light from the sample which is compiled as data for subsequent processing. Figure 1 shows a display on a video display unit ('VDU') incorporated in a system to enable a user to initiate use of the Color Call (RTM) software.

Having received LAB values of a sample software marketed under the name COLOUR CALL (RTM) is used to process the information to undertake a wide range of functions. The particular algorithm involved depends on the furnishing material sample from which the reading was taken. The use of an algorithm matched to the material being sampled ensures that good visual matches can be obtained when crossing from one furnishing material to another. This is desirable since different furnishing types reflect light differently. The sampling system of the present invention ensures that colour comparison is done in a way to overcome the differences, on a common basis. The basis of the present invention is to address the problem of achieving an acceptable visual match between two different types of furnishing material substrates.

BRIEF DESCRIPTION OF DRAWINGS

Exemplary embodiments of the invention will now be described with reference to the three accompanying figures showing screen displays on a VDU in a system according to the present invention. MODES FOR CARRYING OUT THE INVENTION

BASIC USER OPTION: SPECTRAL MATCHES FOR CARPETS

In this embodiment 'COLOR CALL' (RTM) software looks for carpet that most closely matches the sample in terms of pure spectral (LAB) match.

1. First the software processes the LAB values and searches an extensive carpet database for all the carpet that most closely matches the sample in overall appearance and calculates and displays two figures against each. The first figure 'Delta E', also referred to as 'ΔE', represents how close the colour of the item found in the search is in overall appearance to the original sample. The second figure 'Delta H', also referred to as 'ΔH', represents how close the sample is in just hue (colour regardless of lightness and intensity) relative to the item found in the search.

2. The software takes this information and filters all the carpet recovered in the search and displays those closest in just hue to the top of the display screen.

3. Where a value of ΔH (representing difference in hue) is under a set threshold the software sub-filters the information again by ΔE representing difference in overall appearance.

4 Finally the software displays different coloured blobs (referred to as 'traffic lights') against the or each displayed carpet to denote how close the matches are. For example: an excellent match is shown 'green'; a good match is shown 'amber' and a fair to poor match is shown 'red'.

Figure 2 shows a display on the VDU of Figure 1 with, as background, a list of recovered available carpets with the 'traffic light' symbols shown in the right hand column.

In order to allow for the different way that various furnishing types reflect light the 'traffic light' tolerances are separately configured to allow for example much higher tolerances on matching carpet to paint, wallpaper and ceramic tile samples than matching carpet to curtain, upholstery and carpet samples. Figure 2, referred to above in Paragraph 4, shows superimposed on the carpet listing. Using standard CIELAB or CMC ΔE sorting has limitations when trying to match between different substrates because of the very different way in which they reflect light, particularly paint and carpet. If one attempts to use similar tolerances for all types of furnishing material poor results are achieved. Each furnishing material type needs its own set of tolerances and this has been the subject of much development work in the course of progressing the present invention.

Whereas current colour quality control software can find a close match none of the software prior to the present invention has been able to find, for example, contrasts, harmonies, and neutral complements. Up top this point existing systems are effectively 'passive' systems adapted to provide colour matching within certain commercial tolerances. The present invention provides a 'smart' system which can be used to provide for the interpretation of a user's requirements and for their display to a far greater extent than has been possible heretofore. A users requirements can be met by making use of algorithms incorporating recognised interior design principles (in practice usually involving a substantial amount of human input and knowledge). In this way even a user having little or no knowledge of the use of colour to achieve a high level of finished design in a given context.

A major feature of the present invention is the way matches are sorted. Instead of using straight ΔE or Δ H searches the invention makes use of a combination of these in a specific order along with a set of tolerances for each furnishing type to find not just the best spectral match but the one that is the best visually. This specific method resolves the problem that different furnishing types reflect light differently and therefore if you scan one furnishing type to get the best match in another using a standard method such as just a CIELAB ΔE search, software will tend to return very poor visual results. A good illustration of this is scanning carpet to find a paint match. The problem here is that gaps between carpet tufts are read by the spectrophotometer as being black. As a consequence the result comes out as being too dark and dull. The 'COLOUR CALL' (RTM) software provides for an excellent match. It can also readily take into account common features such as a wish by a customer to want a lighter paint than then carpet and therefore allows this as a selection option.

The tolerances used for assessing the quality of the match by COLOR CALL software differ from the tolerances used for manufacturing quality controls.

The result of the filtering and sub-filtering processes incorporated in these steps are output so that values displayed at the top of the screen (and with 'green light') are the best matches possible.

Figure 3 shows a setting-up screen for the system referred to in Figures 1 and 2 enabling a particular choice to made from the processes, methods and displays available.

ADVANCED USER OPTION: CHROMATONE MATCHES AND INTERIOR DESIGN FUNCTIONS. Here instead of the software looking directly for the carpet that most closely matches the sample, the software first finds the closest. Chromatone (RTM) match. This allows the user to access a whole series of interior colour relationships and find the carpet that meets each of these criteria. The relevance of using the Chromatone value rather than Pantone, NCS, RAL or any other is that the software remains focussed on looking only for references that are actually made in carpet and not ones that are not relevant. These interior design functions include the following.

1 Tone: this allows the user to move up and down both the lightness and darkness and the intensity scales.

2 Contrasts: the system uses the circular nature of the Chromatone scale to find the correct contrasts and uses the lightness and intensity references to ensure that the tone remains correct.

3 Harmonies: works on same principle as contrasts.

4 Neutrals: uses the unique geometric layout of each hue tray to find exactly the right neutral complement.

5. The 'close' option allows the user to look all around the specific Chromatone pom in its specific hue tries to find 'close' rather than 'precise' matches.

OTHER OPTIONS AND FEATURES

Show Details- This options displays a comprehensive set of information relating to the specific carpet range.

LOGGING AND REPORTING INFORMATION:

'COLOR CALL' software logs every scan it does including the sample type scanned and its specific spectral data. This is compiled in the form of a report. Also the user can log 'failed matches' which are any instances where the retailer does not have that colour available from within their ranges. This information is also compiled and available in report form.

In what has outlined above it has been assumed that a straight colour match between a sample and one or more available carpets. However the proposed system is capable of providing for other relationships and for comparisons between different materials. Apart from a straight colour match there could be a search to identify a carpet whose colour complements (or is related in some other way to) that of the sample.

Also the previously described exemplary embodiments have involved recovering from a database at least one carpet matching a sample of carpet. However the comparison can be between on the one hand a carpet sample and on the other hand, for example, paint or paper types listed by comparable characteristics on a further database or databases. Typically the comparison could be pursued further by providing a carpet with a wall paper of complementary colour and a paint providing a highlight of appropriate colour for harmonious, or other, relationship with the carpet and paper. PAINT OPTIONS

As an extension of 'CARPET FINDER' ™ , 'CARPET AND PAINT FINDER' ™ gives both carpet matches and paint matches for it. The method adopted is as follows. 1 First the software looks up the closest paint match reference to the sample scanned.

2 Next it goes to a lookup table and finds the closest Chromatone (RTM) value and the carpet ranges which match it.

3 The software then displays both the Dulux (RTM) paint reference and the carpet match.

DATA CAPTURE SOFTWARE

In order for 'COLOR CALL' (RTM) software to work effectively, there needed to be a large database using accurately specified data. This led to the need to develop a capture tool allowing fast data entry, spectral data storage and the attachment of the correct Chromatone (RTM) reference to be attached to the carpet.

The biggest problem with using a scale rather than spectral data for matching is that the best reference cannot always be obtained through colour measurement The only way to ensure that it is totally correct is by a combination of colour measurement and visual assessment. This created a huge dilemma because visual assessment is extremely time consuming. This led to the development of a tool which while allowing for some visual assessment kept it to a minimum.

First the tool functions to find the closest five Chromatone (RTM) values based on the ΔE value. It then filters the results by ΔH. Within certain tolerances the software will narrow down the Chromatone (RTM) references to just one which means this can be dropped straight into the database. Within the next set of tolerances the software recommends visual assessment of two poms of the five. Outside these tolerances the software serves to recommend visual assessment of the three closest poms. Where there is more than one option to be visually assessed a message come up on the screen instructing the user on what to do and having carried out the assessment, allows the user to simply select the reference by one mouse click and drop it into the database with a second mouse click. The second mouse click drops the Chromatone (RTM) value, the spectral data and the LAB data straight into the database. Although the capture tool in this case functions with Chromatone (RTM) values, it can use any search values required such as those involving Pantone (RTM), Dulux (RTM), NCS etc.

INDUSTRIAL APPLICABILITY

The invention provides for a sophisticated system for matching a range of parameters associated with carpet and other fabrics and paper and paints to enable a ready matching of these materials to achieve a desired visual end. The invention also provides for the ready matching of the selection process to a commercial environment such as retailers and stockists of carpets, fabrics, wall paper, and paints

The invention overcomes a colour comparing problem in going from flat surfaces (wallpaper paint and ceramics) to textiles, particularly carpet. The problem here is that flat material, particularly paint and wallpaper samples is quite often far too light to get a sensible carpet match using standard search techniques with the result that light yellows reds and oranges come up as beige or cream matches and blues, lilacs and greens come up as grey or white. The 'COLOUR CALL' (RTM) software overcomes this by specific search and sort techniques combined with the tolerance settings for each furnishing type. Not only does it bring yellow up as yellow match and a blue as a blue but it also takes into account that the customer usually requires darker carpet than the paint and provides for readily finding the right one.

Colour matching across different furnishing types is fundamental feature provided by the present invention. Existing colour matching software is mostly quality control software designed to keep manufactured goods within commercial colour tolerances. Such software for a carpet manufacturer provides for scanning carpet and matching to carpet. For curtain manufacturers matching is to and from curtains. For paint manufacturers are matching to and from paint. It is only in a retail environment that it becomes desirable to match a range of different samples to a particular product range. It is as a result of spending much time developing the present methods that it has resulted in scanning software being developed to operate hardware in a completely different way from use heretofore with the development of the present invention. Whereas current colour quality control software can find a close match none of the software prior to the present invention has been able to find, for example, contrasts, harmonies, and neutral complements. Up top this point existing systems are effectively 'passive' systems adapted to provide colour matching within certain commercial tolerances. The present invention provides a 'smart' system which can be used to provide for the interpretation of a user's requirements and for their display to a far greater extent than has been possible heretofore. A users requirements can be met by making use of algorithms incorporating recognised interior design principles (in practice usually involving a substantial amount of human input and knowledge). In this way even a user having little or no knowledge of the use of colour to achieve a high level of finished design in a given context.

The invention is particularly concerned with colour comparison between materials of different types since existing colour matching systems provide excellent results between similar substrates but fail to provide acceptable results when scanning one substrate (i.e. a paint sample) to find matches with a completely different substrate (i.e. carpet). By using a specific combination of algorithms and allocating a different set of search and sort tolerances to the various furnishing types the invention allows the user to achieve good results between these different furnishings cannot be achieved with existing systems.

Claims

1 A method of colour comparison wherein a sample is scanned by means of a spectrophotometer which provides an output representing the colour of the sample characterised by the steps of:

5 establishing a working value for the sample based on coincidence or any differences up to a predetermined value established during the comparison step for the majority, if not all, of the predetermined number of points; making use of an algorithm depending on the sample of furnishing material involved (that is to say a given algorithm for carpets; a further algorithm for paint, yet another algorithm for wall paper, and so on); and

6 comparing the working value with comparable colour data for a product in order to identify a particular product having a predetermined working colour relationship with the sample.

2 A method as claimed in Claim 1 characterised in that predetermined number of points in the scanning step is sixteen. A method as claimed in any preceding claim characterised in that the product is selected from a group comprising carpet, paint, flooring, fabric, wall covering, ceramic material.

A method as claimed in Claim 4 wherein each furnishing type is allocated its own set of ΔE and ΔH tolerances to allow for the way each sample reflects light.

A method as claimed in any preceding claim characterised in that the step of comparing the working value with comparable colour data for a product is directed to a straight colour match between the sample and the product

A method as claimed in Claim 1, 2 or 3 characterised in that the step of comparing the working value with comparable colour data for a product is directed to a relationship other than a straight colour match (such as a complementary colour relationship) between the sample and the product.

A method as claimed in any preceding claim characterised by the provision of one or more steps wherein comparison can be made between existing colour scales (such as Chromatone (RTM), Dulux (RTM), Pantone (RTM), NCS) or generated colour scales.

A method as claimed in any preceding claim characterised by the inclusion of a further step, following the step of comparing the working value with comparable colour data for a product in order to identify a particular product, of establishing from an existing database information relating to the particular product such as source, location, amount availability, price, delivery time, required associated resource.

A method as claimed in any preceding claim characterised in that the step establishing a working value for the sample based on coincidence or of any differences established during the comparison step for the majority, if not all, of the predetermined number of points there is provided a listing showing working values giving zero difference or differences within a predetermined range or band of ranges in the case of a given sample

A method as hereinbefore described with reference to the accompanying figures.

Apparatus for colour comparison characterised by being adapted for undertaking the method of any preceding claim.