WO1994014052A1

WO1994014052A1 - A method and an apparatus for measuring colour changes

Info

Publication number: WO1994014052A1
Application number: PCT/DK1993/000408
Authority: WO
Inventors: Henrik V. Juhl
Original assignee: Danisco A/S
Priority date: 1992-12-08
Filing date: 1993-12-07
Publication date: 1994-06-23
Also published as: DK147092D0; AU5623494A; EP0667955A1; DK147092A

Abstract

A method of measuring colour changes, such as during a titration, by means of a colour resolution unit (8) decomposing the light of the surface of a sample from at least three colour components, and a device for measuring the signals emitted by each selected spectral interval, as well as by means of a device for evaluating the signals transmitted by the measuring device. According to the invention, a vectorial addition is performed of the signals identifying each selected spectral interval. The vectors matching the signals of the selected spectral intervals are substantially perpendicular to one another and consequently the signals emitted by selected spectral intervals are mutually independent. Then the colour change is measured by a measuring being performed of the vectorial change Id of the colour vector signal IO. As a result a quantitative measurement of the colour change is obtained, said measurement for instance being used for identifying colour changes. In this manner a colour change can be identified automatically and completely automatically.

Description

Title: A method and an apparatus for measuring colour changes

Technical Field

The invention relates to a method of identifying colour changes, prefer¬ ably during a titration, by means of a colour resolution unit decomposing the light of a sample from a number, preferably at least three colour components, and a device for measuring the signals emitted by each selected spectral interval, as well as by means of a device for evaluating the signals transmitted by the measuring device.

Background Art

GB Patent Application No. 2, 183,824 discloses a colour measuring system for identifying colour changes. The system includes three detectors covering their respective identifying area. The signals trans¬ mitted by the detectors are added, said detector signals being con¬ sidered scalars.

Furthermore, DE Offenlegungsschrift No. 3, 134,273 discloses an appar¬ atus for comparing the colour of an article with selected colours. The apparatus comprises a colour TV camera adapted to detect three colour components. The signals matching the colours are compared with speci¬ fic reference values. When one of the colour signals deviates sufficiently from the reference value in question, the article is rejected. This appar¬ atus is, however, unable to provide a quantitative measurement of the total colour difference.

WO 88/01 380 describes a microprocessor controlled titrator. A photo¬ electric measuring probe measures a possible colour change. The re- suiting colour measurement provides a quantitative measurement of the colour change, but all the information contained in the measuring signal is not utilized.

Brief Description of the Invention

The object of the invention is to provide a method, whereby it is poss¬ ible to indicate a quantitative measurement of the total colour change, such as in connection with a titration, in order to automatically identify a colour change spot during an influence.

A method of the above type is according to the invention characterised by performing a vectorial addition of the signals identifying each selected spectral interval, whereby the vectors matching the signals of the selected spectral intervals are substantially perpendicular to one another and thereby mutually independent, and by the colour change being measured by a measuring of the vectorial change of the colour vector signal. The signals emitted by the selected spectral intervals are thereby correctly weighted relative to one another, and a measuring of the vectorial change of the colour vector signal results in a quantitative mea¬ surement of the total colour change.

A method according to the invention may furthermore be characterised by the sum of the vectorially added signals being standardized, and by the colour change being measured by a measuring of the vectorial change of the standardized colour vector signal. As a result the light intensity changes caused by irrelevant conditions cannot influence the measuring result.

Furthermore according to the invention the numeric value of the vectorial change of the colour vector signal may be detected versus a titration, and a colour change corresponding to a percentage of the maximum value of the vectorial change of the colour detector signal may be used for an indication of a colour change. In addition according to the invention an empiric determination may be performed of the percentage of the maximum value of the vectorial change of the colour vector signal, said vectorial change indicating a colour change.

Moreover according to the invention the method may be used for deter¬ mining the alkalinity of a fluid on the basis of the amount of acid which must be added during a titration in order to cause a colour change of approximately 95%.

The invention relates furthermore to an apparatus for carrying out the method according to the invention and for measuring colour changes, said apparatus comprising a colour resolution unit decomposing the light of a samr from a number, preferably at least three colour components, a device .or measuring the signals emitted by each spectral interval selected by the colour resolution unit, as well as by means of a device for evaluating the signals transmitted by the measuring device. This apparatus is characterised by the evaluating device being adapted partly to perform a vectorial addition of the signals emitted by the selected spectral intervals, whereby the vectors matching the signals of the selected spectral intervals are substantially perpendicularto one another, and partly to determine the vectorial change of the colour vector signal. The resulting apparatus is capable of providing a quantitative measur¬ ement of the total colour change.

Furthermore according to the invention, the evaluation device may be adapted to standardize the sum of the vectorially added signals and to indicate the colour change as the vectorial change of the standardized colour vector signal. As a result, the conditions influencing the light intensity cannot influence the measuring result, and accordingly a very accurate identification of the total colour change is obtained. Finally according to the invention a memory may be provided which records the change of the colour vector signal versus a titration, where¬ by a colour change corresponding to a specific percentage of the maxi¬ mum value of the colour change can be used as an indication of a colour change.

Brief Description of the Drawing

The invention is described in greater detail below with reference to the accompanying drawing, in which

Fig. 1 illustrates an apparatus according to the invention for measuring colour changes in a sample by means of an electronic colour resolution unit decomposing the light from the sample into three colour compo¬ nents,

Fig. 2 illustrate the signals matching the individual colour components versus the added amount of acid,

Fig. 3 illustrates how a vector formed by the colour components is changed when acid is added,

Fig. 4 illustrates the numeric value of the vectorial colour change versus the added amount of acid,

Fig. 5 illustrates the difference function of the end portion of the curve of Fig. 4, which serves to indicate whether said portion will change further due to addition of more acid,

Fig. 6 illustrates the end portion of the curve of Fig. 4 for determining the spot identifying a colour change of 95%, and Figs. 7 to 9 illustrate how the method according to the invention can be used for determining COD in waste water.

Best Mode for Carrying out the Invention

The apparatus of Fig. 1 comprises a container 2 for a fluid, the alkalinity of which is to be measured by way of a titration. Fluid samples are removed from the container 2 by means of a measuring head. The fluid samples are transferred to a vessel 4 with a magnetic stirrer 6. A colour resolution unit 8 of the type Coloursensor CS50 (Yamatake-Honeywell) is arranged approximately 5 cm above the surface of the fluid. The colour resolution ur 8 decomposes the light received into three compo¬ nents, viz. a red, a green, and a blue component. The three components are indicated in . . TΠ of a decimal number between 0 and 1 50. In order to avoid a poll ; :;on of the colour resolution unit 8, it is possible to mount a 3 cm long tube thereon, said tube both limiting the visual field and protecting a lens arranged in front thereof. Furthermore, a halogen lamp 10 is arranged above the vessel 4, said lamp illuminating the sur¬ face of the fluid. The acid is dosed by means of a so-called "autobu- rette" 1 2 of the type Dosimat sold by Metrohm.

The colour resolution unit 8 and the autoburette 1 2 communicate through standard buses with an evaluation device in form of a personal computer 14 collecting data concerning colour changes versus the added amount of acid.

A demand exists for developing a calibration-free titration algorithm, whereby it is unnecessary to use standards, references, colours or the like.

In order to estimate the possibility of developing such an algorithm, a titration has been performed by the following method. A sample of 10 ml lemon juice of an alkalinity of about 0.5 has been admixed 30 ml of water, three drops of phenolphthalein indicator, and two drops of antifoaming film. Then the sample was titrated at the same time as a collection and recording was performed of cognate values of colour components and dosed amount of acid.

The result of such a titration appears from Fig. 2. During the first step of the titration the colour components did not change although acid was added to the sample. The colour changed shortly after and gradually as about 1 .5 ml of acid was added. When the sample had been subjected to a complete titration, i.e. after addition of more than 5 ml of acid, a new colour level appeared which was stable and did not change when further acid was added. The colour data concerning a titration can also be illustrated as a vectorial addition of the signals emitted by the selected spectral intervals. The vectors identifying the selected spectral intervals are preferably orthogonal, i.e. the signals emitted by the selected spectral intervals are mutually independent. An illustration of the vectorially added signals, i.e. the colour vector signal, appears from Fig. 3. The initial colour vector signal l₀ and a later colour vector signal I appear, which means that the colour has changed. The colour vector signals are preferably standardized. The difference between the colour vector signals represents then the vectorial colour change l_d. An illustra¬ tion of the numeric value of the vectorial colour change l_d appears from Fig. 4. Based on this illustration it is possible to deduct the same result as from the curves in Fig. 2. The amount of data has, however, been reduced from three colour components to one colour component without the information being discarded. Based on the resulting data the algorithm is created.

a) Before acid is added, the initial colour level is determined by means of the average value of 30 measurements. A measurement of the noise is simultaneously determined and corresponds to the maxi- mum distance between a single colour vector and the average colour vector.

b) The addition of acid is started while a titration is performed at a high speed as long as the distance between the actual measuring and the initial colour vector is less than 3 times the noise.

c) When the distance exceeds this value and the colour has started to change, the titration speed is reduced on the autoburette 12 in order to avoid that the change spot is not passed too quickly. A collection of colour data and dosed amount of acid is simultaneous- ly started.

d) During the last step of the titration, the average distance is con¬ tinuously observed from measuring to measuring. The titration and the collection of data are stopped when the above distance is less than 0.2 times the noise, the latter being considered an indication of the fact that the colour does not change any more, cf . Fig. 5. In other words the titration is stopped at about 5.25 ml of dosed acid.

e) Then the final colour vector is determined as the average value of 30 measurements corresponding to a) .

f) Based on the collected data, the colour change expressed in per¬ centages is determined as the distance between the immediate col¬ our vector and the initial colour vector relative to the distance between the initial colour vector and the final colour vector. The alkalinity is determined on the basis of the filtrated data represent- ing the average value of 5 measurements before and 5 measure¬ ments after the measuring procedure as well as on the basis of the amount of acid to be added in order to obtain a colour change of 95 %. In other words an excessive titration is performed until it is certain that the colour does not change any more. Then the col¬ lected data registered in a memory are reprocessed to provide the amount of acid corresponding to a colour change of 95%. This amount of acid represents then the alkalinity of the sample.

When this algorithm is used it is no longer necessary to know the col¬ ours forming part of the titration. Then a titration changing from for instance yellow into blue can be identified by means of this algorithm. The titration need only follow the procedure shown in Fig. 4.

The detection algorithm and the necessary software for controlling the colour measuring device and the autoburette have been described in the programming language C. A transcript of the programm is attached this specification.

Tests have been performed in order to evaluate the efficiency of the detection algorithm. These tests include a large number of titrations performed on a reservoir juice, where the possible operating limits of the alkalinity measuring device should be determined. The results appear from Table 1 .

The values of a 90% and a 95% colour change, respectively, are indi- cated for each titration.

The standard conditions are: 10 ml reservoir juice 30 ml of water 3 drops of indicator 2 drops of antifoaming oil 1 2 V halogen lamp approximately 200 rpm of the magnetic stirrer. Table 1

Excessive titration due to too much noise in form of "waves" in the vessel.

titrations according to standard conditions are indicated in Table 2. Table 2

The conclusion must be that the developed detection algorithm is very reliable as it has not been possible to trick it even by way of a very rough treatment, such as by addition of a blue colour before the titration or by performing the titration under varying light conditions. Based on the variation coefficient the accuracy is calculated to be about 1.5%, which is considered a satisfying accuracy.

Without involving substantial modifications, the algorithm can be used in connection with other colorimetric titrations.

Figs. 7 to 9 illustrate how the method according to the invention can be used for determining COD in waste water. When determining COD in waste water, the analysis is terminated by a titration of a sample con¬ taining an unknown amount of potassium dichromat. An iron(ll) solution is used as titrant. A ferroindicator is used as indicator.

The titration process is illustrated in Fig. 7 showing that the colour expressed by the three colour components red, green, and blue changes continuously during the titration and very quickly at the change spot.

Fig. 8 shows the length of the vector l_d and the derived function of first order concerning dosed amount of titrant versus the dosed amount of titrant. The colour change of the shown titration has been manually determined to 9.5 ml, which in Fig. 8 corresponds to the derived func¬ tion of first order having reached its maximum value when said colour change occurs.

Claims

Claims.

1. A method of identifying colour changes, preferably during a titration, by means of a colour resolution unit decomposing the light of a sample from a number, preferably at least three colour components, and a device for measuring the signals emitted by each selected spectral interval, as well as by means of a device for evaluating the signals trans¬ mitted by the measuring device, c ha racte r ised by performing a vectorial addition of the signals identifying each selected spectral interval, whereby the vectors matching the signals of the selected spec- tral intervals are substantially perpendicular to one another and thereby mutually independent, and by the colour change being measured by a measuring of the vectorial change of the colour vector signal.

2. A method as claimed in claim 1, c h a r a c t e r i s e d by the sum of the vectorially added signals being standardized, and by the colour change being measured by a measuring of the vectorial change of the standardized colour vector signal.

3. A method as claimed in claim 1 or 2, c h a r a c t e r is e d by the numeric value of the vectorial change of the colour vector signal being identified versus a titration, and by a colour change corresponding to a percentage of the maximum value of the vectorial change of the colour detector signal being used for an indication of a colour change.

4. A method as claimed in one or more of the preceding claims, c h a r a c t e ri s e d by performing an empiric determination of the percentage of the maximum value of the vectorial change of the colour vector signal, said vectorial change indicating a colour change.

5. A method as claimed in one or more of the preceding claims and used for determining the alkalinity of a fluid on the basis of the amount of acid which must be added during a titration in order to cause a colour change of approximately 95%.

6. An apparatus for carrying out the method as claimed in one or more of the preceding claims for measuring colour changes, and com- prising a colour resolution unit decomposing the light of a sample from a number, preferably at least three colour components, a device for measuring the signals emitted by each spectral interval selected by the colour resolution unit, as well as by means of a device for evaluating the signals transmitted by the measuring device, characterised by the evaluating device (14) being adapted partly to perform a vectorial addition of the signals emitted by the selected spectral intervals, where¬ by the vectors matching the signals of the selected spectral intervals are substantially perpendicular to one another, and partly to determine the vectorial change of the colour vector signal.

7. An apparatus as claimed in claim 6, characterised by the evaluating device (14) being adapted to standardize the sum of the vectorially added signals and to indicate the colour change as the vectorial change of the standardized colour vector signal.

8. An apparatus as claimed in claim 6 or 7, characterised by a memory for recording the change of the colour vector signal versus a titration, whereby a colour change corresponding to a specific percen¬ tage of the maximum value of the colour change can be used as an indi¬ cation of a colour change.