DE3940158C2 - Procedure for the objective determination of color perception - Google Patents

Description

The normal human sense of sight makes a known difference Colors like red, green, blue and yellow. But whether that Statement "I see yellow" the same emp in all people finding matches is an undecidable question. Designated the vast majority of subjects agree a first object as "red", a second object stood as "green" and a third item as "blue", so can at least be concluded from an analogous feeling. Accordingly, the scientific color theory is based on the we based on consistent statements of a variety of People.

Also the detection of a defective color, for example with the help of the nail anomaloscope, is based on color statements of the subjects determined. A practicable method without ver bale statements a lack of color in humans delivery is not known.

It is the case for animals that are unable to make a color statement extremely difficult to make a statement about their color perception do.  

For example, Nagel (Zentralblatt für Physiolo gie, June 29, 1907, pp. 205 and 206) of an attempt to Poodle on the retrieval of certain items, with To train names of the named color. The to be retrieved Objects were in numerous brightness levels selected the colors blue or red, so that both red like blue in light, medium and dark shades were there. Against deceptions by the sense of smell various measures taken. Nagel came because of his Try to find out that dogs see colors. Bernhard Grzimek (Journal of Animal Psychology, 1952, No. 9, p. 23) however, came to the conclusion that dogs are color blind. In D. Eisfeld presented another investigation (Investigations about the color vision of some wild canids, special print the Journal of Scientific Zoology, Vol. 1976, Issue 3/4, 1966, pp. 177 to 225), which is the method of Food dressage served, stating that coyotes, jackals and ice cream only reacted to brightness. Only with the wolf he thinks he can prove a color vision. H. Autrum (Color vision of the vertebrates, Tabulae Biologicae, Vol. XXII, 3/2, 1958, pp. 33 to 42) using the example of the domestic cat the great difficulties that come with a conclusive Evidence of color vision in animals are related. With new investigations using problematic methods such as Günter found dressage (J. comp. Physiol. Psychol. 1954, 169), Meyer, Miles & Ratoosh (J. Neurophysiol., 1954, 289), Rücker (Zool. Contribution N.F., 1957, p. 25) no color-specific reac ions; they declared the cat color-blind. Buchholtz (Z. Tierpsych. 1952, 462), on the other hand, combined the play instinct of Cat with the dressage and writes the cat a well wrapped up color sense.

The different results that the authors achieved genes are indeed based on the methodological difficulty To move animals to a "statement". Hartenstein (The Dreidi dimensionality of the colors, proven for the true color  tehmung der Hundes, Die Farbe, 1956, Issue 3/6, pp. 153 to 175) trained dachshunds in such a way that in the experiments presented a large number of colored lights to the dogs that they should sort of sort. With some of the Lights were made known to them during dressage, namely so that the dogs noticed that the dressage lights belonged two different groups, the "Dressurunbunts" and the Dressage colors. The color stimuli used in the experiments were that is produced by filtered incandescent light.

Eisfeld, who worked with shy wild animals, used his dressage training two green-painted wooden boxes of 19 × 14 × 7 cm, which had a hinged lid at the top, that of the Animals could be lifted up with the snout. In the Wooden boxes could be fed. Over the rear A frame was attached to the edge of the boxes in the signalta were inserted. These signal boards were after the DIN color chart (DIN 6164, March 1960 edition) verified. Since the Experiments were carried out outdoors, the signal boards were exposed to different lighting. One animal had two Can distinguish panels by color. This was the case if the dressage on a signal board against an equally bright different colored succeeded. The difficulty was the same Achieve brightness since ice field has nothing to do with the spectral Sensitivity of canine eyes was known. Eisfeld turned thus a method like the one almost 60 years ago by Samojloff and Pheophilaktowa ("On Color Perception with dogs ", Zentralblatt für Physiologie, 1907, p. 133 bis 139) was applied. However, the last noticed mentioned authors that one can not be sure whether the Colors that appear to be equally bright to people for that too animals are equally bright.

Overall, the results obtained by dressage are as problematic to view because it cannot be ruled out that the experimental animals not on colors, but on other opti cal characteristics, e.g. B. Different surface quality have been trained by boxes or paper.  

Other methods have therefore already been proposed those of whom the photoactic reactions, the adaptation the body color to the color of the environment, the partial Er fatigue of the eye due to light of certain wavelength, which in stinctive attachment to objects of a certain color (bold, Tabulae biol. 4, 1927, p. 376), the optomotor reaction heterochromatic flickering and simultaneous color contrast should be mentioned. With the optomotor reactions are on a rotatable cylinder, in the center of which the test animal is, alternately colored with gray stripes. The brightness of the gray is variable abel. There is no grayscale at which the test animal no longer reacts to the rotation of the cylinder Color quality, not brightness, determine the reaction. In heterochromatic flicker, the streams of action become of the eye.

Most of the methods mentioned above do not clear conclusions about color vision because not can be excluded that the brightness or others Influencing factors, for example in the case of rotating cylinders Flicker frequency, determine the behavior of the animals.

In a recent investigation, it was checked again whether dogs recognize colors may or may not (J. Neitz, T. Geist, G, H. Jacobs: Color Vision in the dog, Visual Neuroscience, 1989, No. 3, pp. 119-125). The result of this investigation passed in that dogs have a dichromatic color vision. The dogs were at compared to the tests carried out on a wall, the three color stimulus areas Chen each 2.5 cm in diameter and a mutual distance of 5 cm had, these three surfaces constantly with achromatic light (4800 K, 5th cd / m²) and an additional area with monochromatic light was illuminated. The monochromatic light was used in various experiments changed, and it was the task of the dogs to add the color stimulus area To recognize light. To rule out that the dogs only on brightness differences de addressed, the radiation of monochromatic light was over a large This area was changed at random from one area to another that a normal trichromatic person every monochromatic / achromatic Would judge couple as equally bright. The dogs could be different monochrome distinguish table lights from achromatic light, with the exception of mono chromatic light at 480 nm. Also in this experiment, the dogs trained beforehand and spurred on by reward so that subjective influences do not could be excluded.

A reaction in vertebrates that is largely free of subjective influences is the pupil contraction. In a recent study, Hammond and Mouat ("The relationship between feline pupil size and luminance", Experimental Brain Research, 1985, pp. 485 to 490) confirmed older measurements, according to which the pupil size in cats was approximately linear with the logarithm the luminance decreases. The pupil areas are then reduced from 120 mm² at 10 ^-1 cd / m² to approx. 10 mm² at 10³ cd / m². The investigations by Hammond and Mouat as well as by all other authors who measured the pupil width or area depending on the brightness or luminance, however, do not allow any conclusions to be drawn about the color vision, because the luminance was measured in sizes that were determined by the Brightness sensitivity curve V (λ) of the human and not of the animal eye is determined. In addition, the use of white sunlight or normal reddish-white lamp light as a luminance generator does not allow any conclusions to be drawn about the color vision anyway.

Engelking ("About the pupil reaction with congenital total color blindness", Klin. Monthly sheets for ophthalmology, 1921, p. 707 to 718; "Comparative studies on the pupil reaction with the congenital total color blindness", monthly sheets for ophthalmology, 1922, p. 177 to 188) examined after Sachs ("The change of the pill values with different colored exposure", eg Psych. And Phys. D. P. 1899, vol. 22, p. 386; "About the influence of colored light on the width of the pupil ", Pflügers Archiv fd ges. Physiol. 1892, Vol. 52, 5. 79;" A method for the objective examination of the sense of color ", v. Graefes Arch. 1893, Vol. 39, 3, p. 108 ) and other authors the behavior of the pupils against different colored constant lights, ie he checked the "pupillomotor valence of the colors". He used the v. Hessian pupiloscope, with which the pupillo motor valence of the individual colors could be qualitatively determined. While the pupillomotor valence in the normal sighted was almost 80% in the yellow area and thus a maximum, the maximum of the totally color blind was green and was only 30%. The results given by Engelking are, however, hardly understandable, since neither the colors or light sources presented were specified, nor are details given about the measurement of the pupil width. The specified light transmission of the equivalent gray Goldberg wedges is meaningless if the light rays themselves remain undefined. When evaluating the pupil size for the light adjustment, it should also be borne in mind that the pupil width can vary a maximum of between 2 and 8 mm in adolescence, so that in this way the luminous flux incident in the eye in a ratio of 1:16 and regulates the adaptation of the visual organ to the respective level of illumination to the same extent (Schober, Das See, Volume II, 3rd ed., 1964, p. 40). However, the differences in luminance that the human eye can recognize behave like 1: 10,000,000. However, this does not necessarily mean that the pupil width only plays a role in approximately 1 millionth of the part of the brightness range that can be detected by the eye. Rather, for example, the pupil diameter changes continuously from approx. 2.5 mm at a luminance of approx. 10 ^-5 cd / m² to 3.8 mm at 10² cd / m² (Alexandridis and Dodt, "Pupillary light reflexes and pupil width of a rod monochromatin", Albrecht v. Graefes Arch. Klin.exp.Opthal. 173, 1967, pp. 153 to 161, Fig. 3). This means that the pupil not only blocks the high brightness during the adaptation process and takes over the rest of the adaptation of the so-called visual purple, but that there is a complicated interplay between pupil and other adaptation. Klinker and Spangenberg even want to have found that the largest pupil widths occur at the time of the sun's highest point, the lowest at the time of the sun's low point, which in their view is due to an increase in the sympathetic tone with increasing position of the sun (Z. ges. Hyg. 31, 1985, Issue 2, pp. 88 to 90).

In the essay by Niles Roth "Optometric implications of pupillometry", Journal of the American Optometric Association, Vol. 50, 1979, No. 6, pp. 727-729 are under Different, independent uses of pupillometry shown. For example, it is mentioned that with modern pupillometric Methods objective data about a child's color perception system that is cannot express about colors, can be preserved. From use spectral light sources is out of the question.

In another known method, however, the pupil reaction is made using tested by monochromatic light (Young and Alpern: Pupil responses to foveal exchange of monochromatic lights, J. Opt. Soc. Am., Vol. 70, No. 6th of June 1980, pages 697-706) and measured by means of an infrared television pupillometer. It was found that there was a change from the same bright light At the wavelengths, the pupils contracted when the wavelengths were within a certain range, e.g. B. with a change of 560 nm 498 nm. This procedure is for determining whether there is color vision or not but not particularly suitable, because on the one hand when changing from 560 nm to 650 nm no pupil change occurs and secondly the pupil reaction very much can be observed briefly and only for 50 ms to 500 ms. Hereby the colors can not be seen in the adapted state of the eye.

Finally, it is also known to have objective examinations of color vision ability to carry out with the help of an electroretinogram (SU 1 301 399 A1). Here red, green and blue light stimuli are generated on the cornea with an energy of 0.138; 0.143 and 0.07 mkJ / cm presented. A ratio of an electroretino gram amplitude to a stimulus of different colors within the limits of 0.72-1.4 shows normal color vision. If a change in the ratio of Curve amplitude to stimulus for one color to curve amplitude of stimulus for one other color is less than 0.72 or more than 1.4, there is damage to color vision with this color. The acceptance of electroretinograms is developing very complex.

The invention is therefore based on the object of a relatively simple method specify an objective statement about the color suitability of a subject in the adapted state, whether human or animal, is possible.

This object is achieved according to the features of claim 1. of the subsidiary claim 3 or of the subsidiary Claim 5, ge solves.  

The advantage achieved with the invention is in particular in that on the verbal statement of a person or on the dressage of an animal can be dispensed with to an off say about the color efficiency.

Based on the figures described below, the inven explained in more detail. It shows:

Fig. 1 the known light sensitivity curve of the dark and light adapted human eye;

Figure 2 shows the pupil size of the human eye as a function of the brightness according to Reeves.

Figure 3 shows the pupil diameter of the human eye in depen dependence of the brightness after Flamant.

Fig. 4 shows the pupil width depending on the Lichtwel len length according to Wagman and Gullberg;

Fig. 5 shows the pupil width depending on the Lichtwel lenlänge after Bouma.

In Fig. 1, the spectral light sensitivity curve V (λ) for the brightly adapted human eye and the spectra le light sensitivity curve V '(λ) for the darkly adapted human eye are shown. The two curves are generally known (see O. Reeb, Basis of Photometry 1962, p. 45) and have been defined internationally by the Commission Internationale de l'Eclairage since 1924. They essentially state that optical radiation, which is emitted with the same physical power, is brightest in the range of about 555 nm with bright adaptation and below about 420 nm or above 700 nm are no longer visible.

In the dark-adapted eye, the highest sensitivity occurs at around 500 nm. The brightness of points A and B or C and D are perceived as identical, although they are each assigned different wavelengths. This means that a cyan of approx. 515 nm of a yellow orange of 605 nm is only distinguished by the color tone, but not by the brightness, if both colors are emitted with the same physical power. In experiments on animals, which are based on the dressage method, it is almost always overlooked that the animals u. U. not based on the color, but on the brightness. Although some authors have recognized this source of error and tried to eliminate it with colors of the same brightness, they have regularly assumed the sensitivity of the human eye to light, as shown in FIG. 1. However, this is inadmissible because the spectral sensitivity to light of the animal's facial sense should have been taken as a basis. However, the spectral sensitivity to light of animals is not known, and they cannot be determined in the same way as in humans, because animals cannot match the same brightness and cannot make a statement of equality.

However, the question arises as to whether the spectral sensitivity curves allow a statement to be made about the color temperature, since the fact that different light wavelengths produce different brightness impressions does not mean that a color impression is perceived. A certain indication that the V _λ curves of the brightly adapted eye are also related to the color perceptions is that the V _λ curves of color-defective people differ from the V _λ curves of color-able people (cf. Schober, loc. Cit., P. 221, Fig. 131). In addition to the confusion colors according to Rayleigh's equation, the course of the spectral sensitivity curve plays an important role in differentiating between the different types of color defective vision. The reaction of the pupil may allow conclusions to be drawn about color vision. However, as already mentioned, most authors have examined the pupil size in white light.

In FIG. 2, the examination result of P. Reeves (J. Opt. Soc. Amer., 1920, pp 35 to 48) shown, and for the case that one eye is closed, and for the case that both eyes are open. It is not possible to draw any conclusions about color vision from this illustration.

However, Flamant (Revue d'Optique, Vol. 27, 1948, pp. 751 to 758) came to a somewhat different result than Reeves. The values determined by Flamant are shown in FIG. 3.

In FIG. 4, the opening of the pupil as a function of spectral compared to V _'λ curve of dunkeladaptier th eye is illustrated as it has been found by Wagmen and Gullberg (Am. J. Physiol. 1942, p 769-778) . In contrast to the studies by Reeves and Flamant, the dependence of the pupil width on spectral colors and not on white brightness was measured. It can be seen from the illustration that the values obtained on the basis of objective pupil measurements largely coincide with those values which were obtained on the basis of subjective test subjects. In other words, the size of the pupil changes depending on the spectral colors, analogously to the subjective perception of brightness. Abelsdorff (Zeitschrift für Psychologie und Physiologie der Sinnesorgane, Vol. 22, 1900, pp. 81 to 95) had already determined similar values.

H. Bouma (Nature 193, 1962, pp. 690 and 691) came to somewhat different values. According to his investigations, the pupil sensitivity lies between V '(λ) and V (λ), as shown in FIG. 5. From this it can be seen that the pupil reaction is analogous to the sensation of brightness, but not identical to it. The pupil sensitivity - which is shown in FIG. 5 for the brightly adapted eye - has a sensitivity shift in the blue area. So Bouma tends to come to the same conclusion as Wagman and Gullberg.

One can conclude from this that the chopsticks are at the tax pupil size play a role. This is in the Wi the conclusion on the view of Brown and Page (J. Exper. Psychol., 25, 1939, p. 347), according to which only the cones the pupil control leumotorik.

The spectral pupil sensitivity curve would be identical with the spectral brightness sensitivity curve of the dark-adapted human eye, so could from the spectral pupil sensitivity curve hardly any conclusions be drawn to the color vision because it is then in the essentially a brightness-dependent and not a hue-dependent curve acted. However, the curve lies between V ′ (λ) and V (λ), the cones obviously also work with which, according to the general view, ver are answerable.

Under this assumption, an objective conclusion can be drawn about the color vision: If one looks at the wavelengths in Fig. 5, which both lead to the sensitivity 0.2 at A and A ', it can be seen that on the one hand it is a blue spectral color in the 490 nm range and secondly a red spectral color in the 635 nm range. If both spectral colors are emitted with the same physical intensity, they are also perceived as equally bright by the brightly adapted human eye.

Now take the intensity of both light sources by each half back and if you mix both colors additively, so is the overall physical intensity is the same as before a single light source. When comparing sensitivity this additively mixed light source would have to be sensitive value of 0.2. However, it results in another Value, this difference is based on a color impression, because physically has compared to the previous one Nothing changed. Only the has changed  Color impression because the additive mixture is neither the color blue still results in the color red. Of course you can with the spectral color red at approx. 600 nm and the spectral color Move green at approx. 510 nm. The additive mixed color is then a yellow, similar from the eye in terms of hue how a spectral color in the range of 560 nm is evaluated.

If differences in spectral colors to the right and left of the maximum of the spectral brightness sensation have been found in this way, the same idea can be applied to the pupil curve. One selects two spectral colors that produce the same pupil values at the same intensity (points B and B ′ in FIG. 5), reduces their respective intensity by half, mixes these colors additively and then determines whether the additive color mixture corresponds to the same pupil opening leads a spectral color alone as before. If this is not the case, the difference is due to the color impression and not the brightness. Since the pupil size can also be measured in several animal species, a conclusion can be drawn about the color vision in animals.

According to the Young-Helmholtz three-color theory, which has found a brilliant confirmation in the additive-mixing color television, it would be logical if the spectral sensitivity of the eye at A or A 'in Fig. 5 and the corresponding additive Mixed colors did not differ, so that, in turn, no relevant statement about the perception of color could be made. The Young-Helm holtz theory, however, is based on relatively narrow empirical facts and does not allow any conclusions to be drawn about the physiology of color vision. It remains open whether the three basic colors actually correspond to three types of retinal cones, which v. Studnitz wants to have found. These cones should each contain a colorful visual material, each type of cone a different one. Moreover, Young-Helmholtz's theory cannot explain the counter-colors, so that Hering's counter-color or four-color theory, which uses the three counter-colors "blue-yellow", "red-green", "white-black" underlying, can still not be considered outdated. Rather, based on recent studies (De Valois: Behavioral and Electrophysiological Studies of Prima te Vision. In: Contributions to Sensory Physiology, New York, Academy Press 1968), it appears as if the electrical impulses measured on the optic nerves can, behave as it can only be explained with the help of Hering's counter-color or four-color theory.

Due to the complexity of the visual process and lack of one generally accepted color perception theory, which all observe describes eighth phenomena, it is permissible with an Ab softening of the additive light or pupil sensitivity of the individual sensitivities to a color perception shut down.

When comparing the light sensitivity is another Step possible: It is not only the sensitivity e.g. B. at 600 nm and 510 nm with the additive sensitivity ver be compared, but it is also possible to be sensitive comparative use, which in a purely spectra len yellow occurs, the color tone with the additive-red-green yellow matches.

The technical devices with which the described Measurements made are not described as they are known. For example, the pupil sizes can be changed in humans and animals using infrared photography men.  

The spectral brightness sensitivity curves for day and night vision (V _λ , V _λ ') and the pupil curve V _pup shown in FIG. 4 are combined with the term "eye sensitivity curve". In addition, the term "eye sensitivity curve" includes the measured values of an objectively detectable body reaction that can be quantitatively altered by light incidence as a function of the light wave length at constant light intensity.

Claims (7)

1. Method for the objective determination of an optical color perception, characterized by the following steps:

• 1.1. two spectral lights are selected on both sides of the maximum of an eye sensitivity curve (V ′ _λ , V _λ , V _pup ), in which the same physical radiation power also causes the same sensitivity;
• 1.2. the visual sense to be tested is exposed to at least one of the spectral lights;
• 1.3. the sensitivity of the sense of sight in at least one of the two spectral lights is measured;
• 1.4. the intensities of the spectral lights are reduced by half;
• 1.5. the spectral lights are mixed additively;
• 1.6. the visual sense to be tested is exposed to the additively mixed light;
• 1.7. the sensitivity of the sense of sight to the additively mixed light is measured;
• 1.8. the measured sensitivity of the sense of sight in the additively mixed light is compared with the measured sensitivity of the sense of sight in at least one of the spectral lights;
• 1.9. if the sensitivity of the visual sense in a spectral light differs from the sensitivity in an additively mixed light, it is concluded that there is a color perception.

2. The method according to claim 1, characterized in that the eye sensitivity curve corresponds to the function pupil width = f (λ) = V _pup . 3. Procedure for the objective determination of an optical color perception, characterized by the following steps:

• 3.1. two spectral lights are selected on both sides of the maximum of an eye sensitivity curve (V ′ _λ , V _λ , V _pup ), in which the same physical radiation power also causes the same sensitivity;
• 3.2. the visual sense to be tested is exposed to at least one of the spectral lights;
• 3.3. the sensitivity of the sense of sight in at least one of the two spectral lights is measured;
• 3.4. the intensities of the spectral lights are reduced by half;
• 3.5. the spectral lights are mixed additively;
• 3.6. the visual sense to be tested is exposed to the additively mixed light;
• 3.7. the sensitivity of the sense of sight to the additively mixed light is measured;
• 3.8. the measured sensitivity of the sense of sight with additively mixed light is compared with the measured sensitivity of the sense of sight with that spectral color which corresponds in color to the additive light, both lights having the same physical radiation power;
• 3.9. if the sensitivity of the visual sense in a spectral light differs from the sensitivity in an additively mixed light, it is concluded that there is a color perception.

4. The method according to claims 1 or 3, characterized in that the Emp Sensitivity of the sense of sight is determined by measuring the pupil size. 5. Method for the objective determination of an optical color perception, characterized by the following steps:

• 5.1. two different spectral lights of the same physical radiation power are determined, which cause the same pupil opening;
• 5.2. the pupil openings when the spectral lights act on the sense of sight are measured;
• 5.3. the intensities of the spectral lights found are halved and additively superimposed;
• 5.4. the pupil width is measured, which arises when the additively superimposed spectral lights act on the sense of sight;
• 5.4. a spectral light type is determined which lies in the spectrum between the two spectral light types mentioned and which has the same pupil width as the additively superimposed spectral light type.

5.5. The physical radiation powers of the added spectral lights with the Radiation power of the determined spectral light type between the two mentioned  Spectral light types compared. 6. The method according to one or more of the preceding claims, characterized characterized that the sensitivity of the sense of sight in subtractive super lights are determined.