Customized filters to address color vision deficiency and methods for designing and fabricating such filters

Abstract

The invention relates to the design, fabrication, structure, and use of filters for addressing Color Vision Deficiency (CVD).

Description

[0001] CUSTOMIZED FILTERS TO ADDRESS COLOR VISION DEFICIENCY AND METHODS FOR DESIGNING AND FABRICATING SUCH FILTERS

[0002] RELATED APPLICATIONS

[0003] [1] This application claims priority to U.S. provisional patent application No. 63 / 683,173 filed on August 14, 2024, entitled DIELECTRIC METASURFACES AS CUSTOMIZED DEVICES TO ADDRESS COLOR VISION DEFICIENCY, which is incorporated by reference herein in its entirety.

[0004] BACKGROUND

[0005] [2] Color Vision Deficiency (CVD) affects about 300 million people worldwide. Persons with CVD report frustration with daily activities such as picking ripe fruits and determining doneness of meat. They also experience limitations in professional fields such as aviation, medicine, and the military. The most common issue for persons with CVD is when colors, which are easily distinguishable to persons with normal vision (NV), are confusing, and critical information stored in color is lost to them.

[0006] [3] The color of an object as perceived by the human eye is a function of at least the source of light illuminating the object, the reflectance properties of the object, and the response of the cones in the human’s eye to the illuminated object. Human eyes have three types of cones, commonly referred to as L cones (long wavelength, sometimes referred to as red cones), M cones (medium wavelength, sometimes referred to as green cones), and S cones (short wavelength, sometimes referred to as blue cones). Each of the three types has a sensitivity to light in a different wavelength band. The visible light spectrum spans between about 400 nanometers (nm) and about 700 nm. FIG. 1 is a graph illustrating the spectral sensitivity of the three different types of cones in the typical human eye. As seen in FIG. 1, the L cones are generally sensitive to light between about 500 nm and 650 nm with a peak sensitivity at about 579 nm and trailing off in either direction in a generally bell-shaped curve. The M cones are generally sensitive to light between about 465 nm and 610 nm with a peak sensitivity at about 540 nm (also trailing off in either direction in a generally bell-shaped curve). Finally, the S cones are generally sensitive to light between about 410 nm and 485 nm with a peak sensitivity at about 440 nm and trailing off in either direction in a generally steeper bell- shaped curve than the L and M cones. Thus, as can be seen in FIG. 1, the L and M cones have substantial overlap in their spectra of sensitivity, whereas the spectral band of sensitivity of S type cones is spaced relatively far away from and have very little overlap with the other two types of cones.

[0007] [4] CVD usually results from one (or more) of the L, M, and / or S cone sensitivity band(s) being shifted from the normal range. The type of CVD is determined by which cone is shifted, and the severity is determined by how much the cone is shifted. Most persons with CVD suffer from deuteranomaly, a red-green color blindness, which arises when the spectral sensitivity of the M cones, or medium / green wavelength photoreceptors, is shifted. The spectral shift generally can vary between 1 nm (mild deuteranomaly) to 20 nm (extreme deuteranomaly, or dichromacy known as deuteranopia). For instance, the dashed line in FIG. 1 illustrates the sensitivity band of a person with CVD caused by a shift of the sensitivity band of their M cones to the right by about 10 nm. A person with a 10 nm shifted M cone response is said to have a moderate case of red-green CVD (and more specifically deuteranomaly red-green CVD of 10 nm - DIO CVD for short). Commonly, a person with a red- green CVD has M cones shifted to the right (toward longer wavelengths) by between about 1 - 20 nm or L cones shifted to the left by about 1 - 20 nm. Red-green CVD accounts for about 96% of cases of CVD and results in increased difficulty in distinguishing red and green colors from each other. Red-green CVD generally comes in four different forms, namely, (1) the aforementioned deuteranomaly, which is a red-shift in the M cones, (2) deuteranopia (which is a lack of M cones), (3) protanomaly, which is a blue-shift in L cones, and (4) protanopia, which is a lack of L cones.

[0008] [5] It is also possible to have blue-yellow CVD, in which the S cone spectral tuning is shifted to the right from the normal spectrum (called tritanomaly).

[0009] [6] Unfortunately, current solutions to improve the ability of CVD patients to distinguish colors (e.g., gold nanoparticle-based filters) are deficient in several areas including tunability, customizability, precision, and efficacy. Previous research has attempted to solve CVD using tinted filters, dyes, contact lenses, and eyeglasses [2], These devices all share the same underlying principle of using band- stop filters to filter the spectral overlap between the cones. They generally have limited efficacy due to inability to precisely tailor the filter spectrum and the use of heuristic design methods resulting in non-optimal CVD correction performance.

[0010] [7] Existing filters using gold nanoparticles suffer from a lack of precision due to the inability to create filters with sufficiently narrow pass bands and / or stop bands. SUMMARY

[0011] [8] In one example embodiment, a method of designing a filter for CVD comprises converting cone responses for Normal Vision (NV) to a color coordinate system, converting cone responses of a person with Color Vision Deficiency (CVD) to the color coordinate system, selecting at least one coordinate of the color coordinate system, for at least one wavelength for the at least one coordinate, determine a ratio between a magnitude for the coordinate for NV and the magnitude for the corresponding coordinate for the person with CVD, and designing an attenuation profile that, when combined with the at least one CVD coordinate at the at least one wavelength, will cause the coordinate of the combination to be closer to the corresponding NV coordinate than the corresponding coordinate of the person with CVD.

[0012] [9] In another example embodiment, a method of designing a filter for CVD comprises converting cone responses for Normal Vision (NV) to the brain color coordinate system (BCCS) for a predetermined spectrum, converting cone responses of a person with CVD to BCCS for the predetermined spectrum, converting the coordinates in the brain color coordinate system for NV to CIELAB color coordinate space, converting the coordinates in the brain color coordinate system for the person with CVD to CIELAB color coordinate space, calculating the distance in CIELAB coordinate space between the coordinates for NV and the corresponding coordinates for the person with CVD over the predetermined spectrum, and determining a scaling factor over the predetermined spectrum that minimizes the difference between the NV coordinates and the coordinates for the person with CVD over the predetermined spectrum.

[0013]

[0010] In yet another embodiment, an optical filter for affecting color vision comprises a passband between about 400 nm and about 550 nm, an attenuation notch between about 550 nm and 585 nm, and a pass band between about 590 nm and 700 nm.

[0014] BRIEF DESCRIPTION OF THE DRAWINGS

[0015]

[0011] The accompanying drawings, which are incorporated in and constitute a part of this specification, show certain aspects of the subject matter disclosed herein and, together with the description, help explain some of the principles associated with the disclosed implementations. In the drawings,

[0012] FIG. 1 is a graph illustrating the spectral sensitivity of the three different types of cones in the typical human eye;

[0016]

[0013] FIG. 2 is a graph showing a mapping of the cone response functions of FIG. 1 into the brain color coordinate space;

[0017]

[0014] FIG. 3 is a modified version of FIG.2 in which the wavelength spectrum has been color coded to show the portions of the spectrum where passive filtering can and cannot be used to bring the CVD person’s perceived color closer to that of a NV person with respect to the RG coordinate;

[0018]

[0015] FIG. 4 is a graph showing the transmission characteristics of a passive filter for correcting DIO CVD in accordance with a first embodiment;

[0019]

[0016] FIG. 5 is a graph showing the transmission characteristics of a passive filter for correcting DIO CVD in accordance with a second embodiment;

[0020]

[0017] FIG. 6 is a graph showing the transmission characteristics of a passive filter for correcting DIO CVD in accordance with a third embodiment;

[0021]

[0018] FIG. 7 is a graph similar to FIG. 2, but wherein the RG plot for the NV has been ratiometrically scaled down so that the unaltered RG sensitivity plot for DIO CVD is greater in magnitude than the scaled-down RG plot for the NV person over almost the entire spectrum;

[0022]

[0019] FIG. 8 is a graph showing the ratio between the unsealed CVD RG plot and the scaled- down NV RG plot from FIG. 7 along with the attenuation profile of a filter to reduce the difference between the RG sensitivity values for CVD and the scaled-down NV RG values in accordance with another embodiment;

[0023]

[0020] FIG. 9 is a graph similar to FIG. 8, but further illustrating the YB coordinate plot and showing the attenuation profile of a filter in accordance with a third embodiment taking into account the YB coordinate plot data;

[0024]

[0021] FIG. 10 is a flowchart of a method for designing a passive filter in accordance with an embodiment;

[0025]

[0022] FIG. 11 shows the attenuation profile of a filter designed to minimize color difference and hue difference perceived by an NV person and the exemplary DIO CVD person when isoluminant, monochromatic light is incident;

[0026]

[0023] FIG. 12 illustrates the coordinates of the Munsell color space;

[0024] FIG. 13 shows the attenuation profiles of two filters designed to minimize color difference and hue difference perceived by an NV person versus a DIO CVD person with and without standard illumination;

[0027]

[0025] FIG. 14 is a flowchart of a method for designing a passive filter in accordance with an embodiment;

[0028]

[0026] FIG. 15 is a graph showing the transmission profiles of various filters discussed hereinabove in one graph for comparison purposes;

[0029]

[0027] FIG 16 is a graph showing the color difference between two consecutive hues when the filters represented in FIG. 15 are applied in comparison to unaided CVD and NV;

[0030]

[0028] FIG 17 is a graph showing the color difference between NV and CVD for a person with DIO CVD in four different scenarios; and

[0031]

[0029] FIG 18 is a block diagram of the components of a computer system that may be used to implement the various techniques described herein.

[0032] DETAILED DESCRIPTION

[0033]

[0030] The aforementioned opposing color channel model of color vision is fairly simple and comprises a spectral integration over well-characterized cone sensitivities, followed by a linear transformation that accounts for different gain and the differential perception of each type of cone. A challenge arises because of the large number of possible spectra that can be incident onto the eyes and result in the same color being perceived. The spectral composition of the incident light on the human eye is a convolution of the intrinsic reflectance of the surface, the extrinsic light source falling on the surface, and any spectral filters between the surface and the eyes. For digital displays, on the other hand, the surface reflectance is substituted with the display spectra, which is the superposition of generally three primary colors. As noted above, color vision deficiency (or CVD) arises when the photoreceptor, or cone cell, sensitivities are slightly shifted. Because the brain relies on differential signals from the eyes, a slight shift in a single cone scrambles the color pointers in the brain and morphs the color gamut available for perception.

[0034]

[0031] The following discussion will focus on the design and structure of a filter for a person with deuteranomaly type CVD of medium severity (e.g., a 10 nm red-shift in the green cone sensitivity), as this is the most common type of CVD. However, this is only exemplary. The methods disclosed herein can be adapted to any type and severity of CVD. The goal in each case is to design and manufacture a passive filter that can be incorporated into eyeglasses, contact lenses, windshields, camera lenses, window coatings, monocles, computer screen covers, augmented / virtual reality headsets, etc. to improve the ability of a person with any form of CVD in both distinguishing particular colors that they would otherwise have difficulty distinguishing from each other as well as causing such colors to appear more natural to them (i.e., more like a person with NV would perceive them).

[0035]

[0032] In some embodiments, particularly embodiments for computer and telephone display screens and the like, the filter need not be a physical filter at all, but may be software that applies a customized transformation (e.g., convolution) on the signals output to a display device, such as a computer screen or telephone screen, to change the colors generated on the screen in real time as needed to aid the particular CVD afflicted user (essentially, a digital filter). This includes real-time active digital color correction in augmented / virtual reality headsets and device (phone / computer) screens.

[0036]

[0033] In yet other embodiments, these filters can be used in industrial applications (either by physical, software, and / or hardware), such as color sorting (fruit selection machinery, color segmentation in nature), recoloring algorithms for virtual spaces, etc. Particularly, a filter can be applied that enhances the perceived difference between any two (or more) particular colors that need to be distinguished for a particular industrial (or other) application.

[0037]

[0034] In an embodiment of a method for determining the spectral parameters for a passive filter for correcting CVD and manufacturing such a filter, an initial step is to select a color space that has the same encoding of colors for CVD and NV.

[0038]

[0035] There are many different ways (e.g., coordinate systems) to define the color space for visible light. In fact, the human brain can be considered to use one such color space, herein referred to as “the brain color coordinate space” (BCCS). Specifically, according to the opposing color theory (see Ingling Jr, Carl R., and Brian Huong-Peng Tsou, ’’Orthogonal combination of the three visual channels." Vision research 17.9 (1977 ): 1075- 1082), the L, M and S values are linearly combined (Ingling-Tsou matrix) in the brain to form a new, three coordinate system. In essence, the brain has a lookup table for the following three values:

[0039] (1) brightness, (WS) (2) how yellow or blue the color is, (YB), and

[0040] (3) how red or green the color is, (RG).

[0041]

[0036] As this is a linear space, the values for any spectrum will be the sum of the individual values because the spectrum is a superposition of delta function spectra at every wavelength. This color space is the same for persons with normal vision (NV) and CVD.

[0042]

[0037] These three coordinates encode spectral responses of the photoreceptors (i.e., the cones) in arbitrary units (a.u.) since the values are normalized . More specifically, the WS coordinate defines luminosity (how bright the color is) and specifically encodes the relative gain between the photoreceptors, i.e., the green or medium wavelength sensitive photoreceptors (M cones) have higher gain than the long and short wavelength sensitive photoreceptors (L and S cones). The RG coordinate defines the red-green axis, where the activation of the long and medium wavelength sensitive cones (the red and green photoreceptors) is compared to give a color between red and green. Finally, the YB coordinate is a comparison of the activation of the short wavelength sensitive (or blue) photoreceptor (the S cones) to the weighted activation of the long and short wavelength sensitive photoreceptors (the L and M cones).

[0043]

[0038] FIG. 2 shows a mapping of the cone response functions of FIG. 1 into the BCCS. The abscissa is wavelength in nms and the ordinate is normalized sensitivity to light at that wavelength. The solid black line shows the plot for the WS (brightness) coordinate for a person with NV, the solid blue line shows the plot for the YB (how yellow or blue the color is) coordinate for a person with NV, the solid orange line shows the plot for the RG (how red or green the color is) coordinate for a person with NV, the dashed black line shows the plot for the WS coordinate for a person with DIO CVD, the dashed blue line shows the plot for the YB coordinate for a person with DIO CVD, and the dashed orange line shows the plot for the RG coordinate for a person with DIO CVD.

[0044]

[0039] A passive spectral transmission filter changes the incident spectra entering the eye by controlling how each wavelength of light is differentially transmitted through it. A passive filter has no amplification mechanisms, and, thus, can only relatively absorb some of the incident light in certain bandwidths. That is, it can only decrease (or leave unchanged) the amount of light transmitted through it relative to the incident light in any bandwidth. It cannot increase it. Thus, the output from the filter (i.e., the portion of the incident light that is transmitted through the filter) always has lower cumulative power density than the input (i.e., the incident light), and the transmitted light will always have lower luminance than the incident input light.

[0045]

[0040] The fraction of the incident light that passes through the filter is determined by the scaling factor and is a function of the wavelength of light. The only design parameter in a passive filter is scaling factor as a function of wavelength (i.e., what percentage of the incident light is transmitted through the filter at each wavelength). Also, by definition, a passive filter is not active, i.e., the scaling factor at each wavelength is not dynamic. That is, it does not change over time.

[0046]

[0041] Thus, after a suitable color space is defined, the next step is to determine the optimal theoretical scaling factor at each wavelength to correct any given individual’s CVD (e.g., make the colors perceived by an individual observing a scene through the filter closer to what a person with NV would perceive).

[0047]

[0042] To design the optimal filter to aid in color identification and distinguishability, the desired relative ratios of attenuation between wavelengths to compensate for the shift in cone sensitivities must be determined for the particular individual with CVD.

[0048]

[0043] Ideally, for a monochromatic source of light, the attenuation factor would be equal to the ratio between (1) the brain color space vector for the incident light in NV and (2) the brain color space vector for the incident light in CVD. These ratios can be determined from the plots in FIG. 2. Particularly, at any given wavelength, the ratio between the absolute value of the solid line and the corresponding (same color) dashed line in FIG. 2 is that ratio. In actuality, this scheme cannot be realized exactly because the NV and CVD vectors are not co-linear, and the scaling factor cannot be different for the different coordinates that make up the vector. Rather, the scaling factor must be the same for all three coordinates [WS, YB, RG],

[0049]

[0044] For CVD of the deuteranomaly type (the most common type of CVD), the red-green coordinate, RG, is the dominant coordinate in the deviation. Thus, it makes sense to scale all three coordinates using the proper scale factor for the RG coordinate in BCCS in order to achieve the most effective scaling to correct the CVD at any given wavelength.

[0050]

[0045] Furthermore, correction for CVD using a passive filter can only be done for wavelengths where the magnitude (absolute value) of the relevant coordinate for a person with CVD is greater than the magnitude of the absolute value of that coordinate for a person with NV (and those two absolute values are of the same polarity). This is because a passive filter can only attenuate light. It cannot amplify light. Hence, a passive filter can bring the CVD person’s perceived color closer to that of a NV person only in portions of the spectrum where that can be achieved by reducing the light transmitted through the filter.

[0051]

[0046] FIG. 3 helps illustrate this point. FIG. 3 is a modified version of FIG.2 in which the wavelength spectrum has been color coded to show the portions of the spectrum where passive filtering can and cannot be used to bring the CVD person’s perceived color closer to that of a NV person with respect to the RG coordinate. Note that, the difference between an NV person’s sensitivity and the CVD person’s sensitivity is different for each coordinate, WS, YB, or RG. Thus, the portions of the spectrum in which passive filtering can and cannot be used to bring the CVD person’s perceived color closer to that of a NV person is different for each of the three coordinates. Thus, one of the coordinates should be selected for correction. Since we have already established that the RG coordinate is the dominant coordinate in the deviation between an NV person and our exemplary person with CVD of the deuteranomaly type (the most common type of CVD), we select the RG coordinate for determining where we can use passive filtering to correct CVD.

[0052]

[0047] The purple bands in FIG. 3 are the bands where the magnitude of the CVD person’s RG sensitivity is greater than the magnitude of an NV person’s RG sensitivity (i.e., where passive filtering can be used to correct the CVD person’s perceived color closer to that of a NV person). The grey bands are the bands where the magnitude of the CVD person’s RG sensitivity in those bands of the visible spectrum is less than the magnitude of an NV person’s RG sensitivity (i.e., where passive filtering cannot be used to correct the CVD person’s perceived color closer to that of a NV person). For sake of completeness, the white bands are the portions where the CVD person’s RG sensitivity is of a different polarity (positive / negative) than the NV person’s RG sensitivity. In those bands of the spectrum, it also is not possible to use passive filtering to correct CVD (essentially the math does not work in those portions of the spectrum).

[0053]

[0048] For CVD of the deuteranomaly type (the most common type of CVD), the red-green coordinate, RG, is the dominant coordinate in the deviation. Thus, it makes sense to scale all three coordinates using the proper scale factor for the RG coordinate to achieve the most effective scaling to correct the CVD at any given wavelength.

[0054]

[0049] FIG. 4 is a graph showing the transmission characteristics of a passive filter in accordance with a first embodiment for correcting DIO CVD. In FIG. 4, the abscissa is wavelength and the ordinate represents the ratio in the hrain coordinate space between NV and CVD for this particular type of CVD. The black line shows the ratio between NV and CVD for the WS coordinate, the blue line shows the ratio between NV and CVD for the YB coordinate, and the yellow line shows the ratio between NV and CVD for the RG coordinate. The green line represents the transmission characteristics of a passive filter designed in accordance with a first embodiment for correcting DIO CVD represented by the black, blue and green lines. Note that the filter attenuation at any given point must be between 1 (meaning there is no attenuation) and 0 (meaning full attenuation - no light of that wavelength is allowed through). As can be seen in FIG. 4, this filter essentially is a double notch filter with a first attenuation notch in the blue portion of the spectrum centered at about 480 nm and a second notch in the red-green portion of the spectrum (about 560 nm to about 585 nm, with peak attenuation at about 585 nm, i.e., it acts on the green -red part of the visible spectra).

[0055]

[0050] For red-green color confusion, the shorter wavelengths do not contribute to red-green color confusion as much, so the shorter wavelengths may not need to be scaled. Thus, FIG. 5 is a graph showing the transmission characteristics of a passive filter for correcting CVD of the deuteranomaly type of 10 nm (the same as in FIG. 4) in accordance with a second embodiment in which the first notch is eliminated. As can be seen in FIG. 5, this particular filter is a single notch filter. It is substantially identical to the filter of FIG. 4, but with the first notch (in the blue portion of the spectrum) removed.

[0056]

[0051] FIG. 6 is a similar graph for the same D10 CVD showing the transmission characteristics of a piecewise filter in accordance with a third embodiment. This filter essentially comprises three notches and corrects for all the coordinates that can be scaled in the visible spectrum.

[0057]

[0052] In another embodiment illustrated with the help of FIGS. 7, 8, and 9, one may take the plot for the NV person (for one or more of the coordinates) and ratiometric ally scale it down (while leaving the corresponding plot for the CVD person unchanged) to the point where the sensitivity for the CVD plot is greater than the sensitivity for the scaled-down NV plot over the entire spectrum (or almost the entire spectrum), and then design the filter using the ratios between those two plots at the various wavelengths to determine when and where to apply attenuation.

[0058]

[0053] Fig. 7 illustrates this concept. FIG. 7 shows the same plot of the RG sensitivity in the brain opposing color coordinate space for D10 CVD from FIGS. 2 and 3 in dashed line. It also shows the same plot of the RG sensitivity in the brain opposing color coordinate space for an NV person in solid thin line. A third line (the thick solid line) shows the RG plot for the NV person ratiometrically scaled down such that the RG sensitivity of the DIO CVD person is greater in magnitude than the scaled-down RG plot for the NV person over almost the entire spectrum. Particularly, note that there are still two small wavelength bands (indicated by grey shading in FIG. 7) where the scaled-down RG plot for NV and the RG plot for CVD are of opposite polarity and / or the NV plot still has greater magnitude than the CVD plot (such that passive filtering will not work in those bands).

[0059]

[0054] FIG. 8 is a graph showing, in orange, the ratio between the unsealed DIO CVD RG plot and the scaled-down NV RG plot from FIG. 7, and shows, in green, the attenuation profile of a filter that will minimize or at least reduce the difference between the RG sensitivity values for CVD and the scaled-down NV RG values.

[0060]

[0055] FIG. 9 is similar to FIG. 8, but also takes into account the YB coordinate. Particularly, as in FIG. 8, the ratio between the unsealed DIO CVD RG plot and the scaled-down NV RG plot is shown in orange. The blue line has been added and represents the ratio between the unsealed YB plot of the DIO CVD person and a scaled-down YB plot for an NV person. The green plot shows the attenuation profile of a filter that will minimize or at least reduce the difference between the RG sensitivity values for CVD and the scaled-down NV RG values as well as the difference between the YB sensitivity values for CVD and the scaled-down NV YB values. For completeness, the black plot represents the ratio between the unsealed WS plot of the DIO CVD person and a scaled-down WS plot for an NV person, but is not considered in the design of the filter.

[0061]

[0056] FIG. 10 is a flowchart showing the steps for designing an attenuation filter in accordance with the principles described above.

[0062]

[0057] In step 1010, a color space coordinate system is selected. In the exemplary embodiment described above, the BCCS system was used.

[0063]

[0058] In step 1012, the cone responses for NV and the cone responses for the particular individual with CVD that is to be treated are converted into the selected color space coordinate system.

[0064]

[0059] In step 1014, one or more of the coordinates in the color space are chosen for correction. This selection would normally be based on the particular CVD type and severity of the treated individual. Most likely, the cone type that is most offset from NV will largely dictate which coordinatc(s) should be selected. For instance, assuming BCCS and dcutcranomaly type CVD (rightward shift of the M (green) cones, then it is the red-green coordinate in BCCS that is probably most offset from NV; and should be chosen. On the other hand, if the person being treated has tritanomaly (leftward shift of the S cone response), then, most likely it will be the yellow-blue coordinate in BCCS that will be most offset from NV; and the coordinate that should be chosen for correction. As previously noted, one may select more than one coordinate also.

[0065]

[0060] Next, in step 1016, at each wavelength of interest (which could be the entire visible spectrum, a portion of the entire visible spectrum or several discreet portions of the visible spectrum, for instance), calculate the ratio between the coordinate for an NV person and the coordinate at the same wavelength for the CVD person being treated. This, for instance, will result in a plot such as the orange and blue plots shown in FIG. 4 (orange being the RG plot and blue being the YB plot).

[0066]

[0061] Next, in step 1018, design an attenuation profile that, when incident light passes through a filter with that attenuation profile, the output (the light transmitted through the filter) will be shifted closer to the NV plot than the original CVD plot in the selected color space. In theory, if this is performed based on the ratios for just one of the brain perception functions (e.g., just for the RG ratio), then this attenuation profile would be identical to the RG ratio plot (e.g., the orange plot in FIG. 4, for instance). That is, in theory, the plot of the desired scaling factor over the spectrum of interest may be identical to plot of the ratios for that brain perception function.

[0067]

[0062] Also relating to the issue of precision, it should be understood by those in the related arts that, while the discussion here (and the graphs in the FIGS.) generally speak in terms of continuous ranges of wavelength (i.e., spectra), in actuality, the plots and other data are generally generated from measurements made at discrete wavelengths (and, e.g., interpolating between those data points to generate a smooth plot). The resolution of those data points, e.g., the scaling factor data points used to generate the attenuation profile, may be any reasonable resolution. For example, since the calculations of the ratios and the attenuation profile are just a matter of rote repetition of the same simple mathematical algorithm at each selected wavelength (and assuming that the operation is performed by software, which is extremely efficient at performing rote repetitive mathematics), a resolution as fine as 1 nm may be selected. On the other hand, as a practical matter, any resolution of 10 nm or finer would likely result in a filter that is more than adequate for most practical applications (it is unlikely that a person with even the most acute color vision would be able to tell the difference between a filter designed using 10 nm wavelength resolution and 1 nm resolution).

[0068]

[0063] As should be clear from the discussion above, it may be very difficult or even impossible to attain perfect correspondence between the output of the filter and the NV plot line. However, a corrected plot line that is, on average, closer to the NV plot line than the original CVD plot line should improve the treated person’s CVD symptoms.

[0069]

[0064] There is significant publicly available data as to which colors are most important to perceive correctly. There also is significant publicly available data establishing the most common types and severities of CVD. Accordingly, to the extent that a perfect filter cannot be constructed, it would be advisable to concentrate on maximizing the correction for the most important colors and / or designing attenuation profiles for the most common forms of CVD.

[0070]

[0065] Finally, in step 1020, fabricate the filter (whether a software filter, e.g., for a computer display or a physical translucent sheet) having as close to the designed attenuation profile as possible. Given any particular attenuation profile, techniques for fabricating a filter having that profile are known in the related arts. Nevertheless, additional, innovative techniques for fabricating these filters are discussed further below in this specification.

[0071]

[0066] While equality in the brain color coordinate space per the above-discussed embodiments can be used to determine naturalness, it is not perceptually uniform. Thus, in accordance with alternate embodiments, one may instead transform the cone responses to the CIELAB coordinate system (hereinafter CIELAB) and, instead, calculate the color difference and color contrast in quantifying deviations in perception between NV and CVD persons and whether two colors are distinguishable. While it may be possible to convert directly from cone response to CIELAB for at least NV, it is better to first convert cone response to BCCS, and then from BCCS into CIELAB . Furthermore, to comprehensively study color vision, it is useful to characterize the role of environmental lighting and the choice of display primaries in the confusion of colors for a person with CVD relative to a person with NV.

[0072]

[0067] Thus, in an alternate embodiment, the focus of the filter design may be minimization of the perceptual difference between the color perceived by an NV person and a person with CVD, rather than matching the ratio between the color space vectors. In one such embodiment, the filter is designed to trace the attenuation factors that minimize color difference between NV and CVD for monochromatic, isoluminant sources.

[0073]

[0068] For such embodiments, converting to and using the CIELAB color coordinate system or another perceptually uniform color space, is a better choice than BCCS because the color space must quantitatively emphasize differences in perception. Considering CIELAB first, while it is possible to convert directly from the cone response to CIELAB, it is more common to convert from BCCS to CIELAB, particularly for non-NV vision.

[0074]

[0069] Once in CIELAB, one can calculate the distance in the CIELAB space between the NV plot and the CVD plot at each wavelength in the relevant spectrum. Then, an attenuation filter can be designed that, when applied to the CVD plot, minimizes the distance between the CVD plot and the NV plot. Designing such a filter is a matter of optimization over the entire spectrum of interest, which can be performed by any available software package for optimization, such as the L-BFGS-B algorithm implemented by the function minimize in Python. The inputs would be the distance at each wavelength between the NV plot and the CVD plot and the program would be configured to generate the set of scaling factors that, when applied to the CVD plot, would minimize the average distance of the combined plot (i.e., the combination of the original CVD plot with the scaling factor plot) from the NV plot over the selected spectrum of wavelengths.

[0075]

[0070] FIG. 11 shows the attenuation profile of a filter designed to correct DIO CVD in accordance with the focus on minimization of the perceptual difference between the color perceived an NV person and a person with CVD using the CIELAB color coordinate system.

[0076]

[0071] As previously noted for all of the above filters, there are wavelength bandwidths that cannot be corrected for without active elements because a passive filter can only lower the input into a person’s eyes (effectively decreasing the response to stimuli), whereas, for certain wavelengths, it would be desirable to increase the response to stimuli to achieve the desired result. A workaround is to lower the spectral stimuli uniformly, and then ratiometrically scale the dominant color axis.

[0077]

[0072] In another embodiment, a similar process may be performed to generate a filter attenuation profile focusing on optimizing for all the hues in the Munsell color coordinate space. FIG. 12 helps illustrate the Munsell color coordinate space (hereinafter Munsell). It is a useful color formalism for understanding how these filters affect real colors because it gives an intuitive understanding of color. In the Munsell HSV color space, there are three coordinates, namely, hue (H), saturation (S), and value (V). Hue describes the base pigment and is the radial coordinate around the axis 1201, as illustrated by the ring of hues 1205. Saturation is the horizontal coordinate (i.e., the distance from the axis 1201 as illustrated by the increasingly saturated arc segments 1203) and describes the amount of base pigment. Value describes the amount of white-black pigment in the color and is the vertical coordinate, as illustrated by the gradation from black at the bottom to white at the top of the axis 1201.

[0078]

[0073] FIG. 13 shows the attenuation profile of a filter designed to minimize color difference calculated in CIELAB perceived by an NV person and the exemplary D10 CVD person with and without CIE (International Commission on Illumination) standard illumination D65 (average daylight) for a subset of spectrum of colors in the Munsell color space. The coordinate for the incident light in the dominant hue radial direction is almost preserved between NV and CVD, but the value of the color is severely reduced.

[0079]

[0074] FIG. 14 is a flowchart showing the steps for designing an attenuation filter in accordance with the principles described above.

[0080]

[0075] In step 1410, a color space coordinate system is selected. CIELAB was used in the two particular exemplary embodiments described above.

[0081]

[0076] In step 1412, the cone responses for NV and the cone responses for the particular individual with CVD that is to be treated are converted into the brain coordinate color space.

[0082]

[0077] In step 1414, the BCCS plots for each of NV and CVD are further converted into the perceptually uniform space (here we have chosen CIELAB)).

[0083]

[0078] Next, in step 1416, at each wavelength of interest (which could be the entire visible spectrum, a portion of the entire visible spectrum or several discreet portions of the visible spectrum, for instance), the distance in the selected color space between the coordinate for an NV person and the coordinate at the same wavelength for the CVD person being treated is calculated.

[0084]

[0079] Next, in step 1418, the scaling factor plot is generated that, when applied to the CVD plot, would minimize the average distance of the scaled CVD plot to the NV plot over the relevant spectrum. The process of designing such an attenuation profile is a matter of optimization over the entire spectrum of interest and is well known in the related arts.

[0085]

[0080] Finally, in step 1420, a filter (whether a software filter, e.g., for a computer display or a physical translucent sheet) is fabricated having the designed attenuation profile. As previously noted, techniques for fabricating a filter having any particular attenuation profile are known in the related arts. Nevertheless, additional, innovative techniques for fabricating such filters arc discussed further below in this specification.

[0086]

[0081] A comprehensive color vision test, the Farnsworth-Munsell Hue 100 test, relies on the ability of a person to arrange equally spaced hues of the same value and saturation in the right order. Fortunately, there is a large database of broadband reflective spectra from Munsell paint chips, and one can model the results of the Farnsworth-Munsell Hue 100 test with different filters, illuminations, and CVDs.

[0087]

[0082] FIG. 15 is a graph showing the transmission profiles of the three filters discussed above in one graph for purposes of comparison with each other. Specifically, the purple plot is the single notch filter of FIG. 5, the brown line is the hue optimized filter with no illumination of FIG. 13, and the blue plot is the monochromatic source optimized filter of FIG. 11. It can be seen in FIG. 15 that, while the various filters share some significant similarities (e.g., significant attenuation around 470 nm and around 570-600 nm, there are significant differences between the various designs.

[0088]

[0083] The plots show that, for hue subsets that are difficult for persons with D10 CVD to distinguish (i.e., the red-green portion of the spectrum), all three of the exemplary filter designs represented in FIG. 15 generally increase the color difference perception.

[0089]

[0084] FIG. 16 is a graph of plots showing the difference in perceived color between consecutive colors in the Munsell color palette for the embodiments discussed above and shown in FIG 15. The graph also includes plots for NV and unaided CVD for comparison purposes. The vertical color difference scale is in Standard Normal units. A value greater than 2.0 on the color difference scale is generally considered to be acceptable (i.e., a person with that visual perception at that wavelength has a reasonable ability to distinguish the corresponding color from consecutive colors in the Munsell palette). The gray plot shows the perceived color differentiation for a person with NV. The black plot shows the perceived color differentiation for a person with D10 CVD. The purple plot shows the perceived color differentiation for a person with D10 CVD using the single notch filter of FIG 4. The blue plot shows the perceived color differentiation for a person with D10 CVD using the monochromatic optimized filter of FIG. 11. Finally, the brown plot shows the perceived color differentiation for a person with D10 CVD using the hue optimized filter of FIG. 13.

[0085] Also for comparison purposes, FIG. 17 is a graph showing the color difference between NV and CVD for a person with DIO CVD in four different scenarios, namely, (a) unaided DIO CVD (the black plot), (b) DIO CVD using the single notch filter of FIG. 5, (c) DIO CVD using the monochromatic source optimized filter of FIG. 11, and (d) DIO CVD using the hue optimized filter of FIG. 13. Although this graph seems similar to the graph of FIG. 16, it shows significantly different information from the graph of FIG. 15. Particularly, as previously noted, each plot in FIG. 16 represents the difference in perceived color between consecutive colors in the Munsell color palette. Thus, a higher number in the graph of FIG. 16 generally indicates a better ability to distinguish colors (because it indicates a better ability to distinguish between consecutive colors in the Munsell scale). Each plot in FIG. 17, on the other hand, represents how much different the color perceived by the DIO CVD person is from what is perceived by an NV person. Thus, in FIG. 17, a lower number is generally indicative of a better ability to distinguish colors (because a lower number indicates color perception that is closer to normal (NV) than a higher number).

[0090]

[0086] Hence, as can be seen in FIG. 17, the single notch filter of FIG. 5 and the monochromatic source optimized filter of FIG. 11 offer significant improvement for the DIO CVD person in the red-green area, which is the area most troublesome for DIO CVD persons. The hue optimized filter of FIG. 13, on the other hand, is generally less effective in that area of the spectrum.

[0091]

[0087] While we have used the example of a person with DIO CVD throughout this specification, it should be appreciated that this is merely an example, and that the methods, apparatus, techniques, and steps disclosed herein may be used to design an attenuation profile for a filter to correct for CVD of any particular kind and / or severity. In embodiments, the exact differences between NV and the CVD of any specific individual may be determined (e.g., using the Famsworth-Munsell Hue 100 test), and a filter adapted to correct that particular CVD profile may be designed using the techniques described above, and then a filter may be fabricated with that attenuation profile.

[0092]

[0088] The above-described inventions, tests, and research results have revealed that some wavelengths cause no confusion for DIO CVD (the mid-range of the most common form of CVD). Thus, it may be advisable to use narrowband sources (like LEDs) centered at these wavelengths for critical infrastructure, e.g., traffic lights.

[0089] The above-described inventions, tests, and research also have revealed that there are certain hues that arc not confusing. Thus, it may be advisable to use these particular hues for conveying critical information in the real world.

[0093]

[0090] Once a filter is designed in accordance with the principles described hereinabove, it must be fabricated. As previously noted, for some applications, the filters may have physical structures that the incident light passes through. Such applications may include, for instance, eyeglasses. In other embodiments, such as filters for altering the colors rendered on a computer screen, telephone screen, or augmented / virtual reality headset to aid a CVD person, the filtering may be effected exclusively by software that alters the original / input colors presented by a source (e.g., a web page) according to the desired attenuation profile. Of course, alternately, it is still possible to use a physical filter in such computer screen applications (e.g., positioning a translucent sheet having the desired attenuation profile over the screen).

[0094]

[0091] Various techniques for fabricating physical visible light attenuation filters are well known in the related ails and need not be discussed here. Furthermore, various software techniques for implementing digital visible light attenuation filters are well known in the related arts and need not be discussed herein.

[0095]

[0092] However, as can be seen from the attenuation profiles of the exemplary embodiments above, some such filters call for very narrow attenuation bands. While various existing fabrication technologies can be used to fabricate filters with narrow attenuation bands, nanoparticle metasurfaces present a particularly favorable technique for filtering in the visible light spectrum. Particles at the nanoscale are at the same order of magnitude as visible wavelengths of light. These particles, if fabricated out of dielectric nanoparticles, may very finely tune light to very specific needs [1], This is achieved by altering the shape, size, and periodicity of the nanoparticles laid out in an array on a substrate. One known application of such a nano-particle metasurface is as a device that aids those with an Anomalous Trichromacy (which is a common form of color vision deficiency, CVD). Although there have been numerous attempts at aiding CVD, such as EnChroma glasses, independent experiments have shown that improvements to vision are insignificant [2], Researchers have also shown that nanoparticles (NPs) have the capability of acting as an aid for CVD [3,4]. To date however, the only types of NPs demonstrated have been metal nanoparticles such as Au.

[0093] In some embodiments, there may be provided a novel use of dielectric nanoparticles (which have low loss and arc far more tunable than metal NPs) to improve distinguishability of colors, such as the distinguishability of colors along a confusion line. The confusion line is a straight line through a co-punctal point on a Commission on Illumination (CIE) chromaticity diagram that roughly approximates colors that a specific type of CVD would struggle with [5]. In some embodiments, a customized dielectric NP based metasurface can be used to filter light to aid color vision deficiency by making colors more distinguishable.

[0096]

[0094] A class of metasurfaces (which are composed of periodically arranged silicon nitride nanoparticles on a silicon dioxide surface) has been developed [7], These particles and their periodic spacing are smaller than the wavelength of visible light to ensure the desired precise color-filter effect.

[0097]

[0095] In some embodiments, the filters may be designed using these previously developed methods (see, e.g., [1]) that allows the generation of highly tunable, narrow-bandwidth optical filtering. This effect is the result of the precise cutting of corners of the silicon nitride cubes to result in t-shaped structures. Prior work [1], where these filters operated in reflection mode, showed these filters operating with polarization-insensitive with equilateral square geometries (see S.I. of [1]) and polarization-sensitive with rectangular geometries (see main text of [1]) to achieve color effects with increased saturation and more precise spectral tuning compared to prior methodologies. Accordingly, some embodiments of the filter disclosed herein may utilize the equilateral square shape basis of [1] to operate in transmission mode.

[0098]

[0096] In the generated filter device, light passes through the metasurface color filter and colors which were previously indistinguishable for a patient with color vision deficiency will become distinguishable when they reach the patient’s eye. The hardware applications in glasses, windshields, display or cell phone covers, window coatings, camera lens coatings, and the like may improve the quality of life of patients with color vision deficiency. In contrast to past approaches, the disclosed CVD filter is customizable based on the type and severity of CVD for an individual patient.

[0099]

[0097] In some implementations, some of the aspects disclosed herein, such as the methods for designing filter attenuation profiles to address particular CVD cases discussed above (e.g., see the flowcharts of FIGS. 10 and 13) as well as the software-implemented filters for computer screens discussed above may be implemented in a computing system, such as a computing system 500 as FIG. 18. For example, the system may be used to process the instructions for the noted algorithmically-optimized software solution as well as other aspects disclosed herein. For example, the system 500 may include a processor 510, a memory 520, a storage device 530, and input / output device 540. The processor 510, the memory 520, the storage device 530, and the input / output device 540 can be interconnected via a system bus 550. In some implementations of the current subject matter, the processor 510 can be a single-threaded processor. Alternately, the processor 510 can be a multi-threaded processor. The processor 510 is capable of processing instructions stored in the memory 520 and / or on the storage device 530 to display graphical information for a user interface provided via the input / output device 540. The memory 520 is a computer readable medium such as volatile or non-volatile that stores information within the computing system 500. The memory 520 can store data structures, instructions, and / or the like. The storage device 530 is capable of providing persistent storage for the computing system 500. The storage device 530 can be a floppy disk device, a hard disk device, an optical disk device, or a tape device, or other suitable persistent storage means. The input / output device 540 provides input / output operations for the computing system 500. In some implementations of the current subject matter, the input / output device 540 includes a keyboard and / or pointing device. In various implementations, the input / output device 540 includes a display unit for displaying graphical user interfaces. According to some implementations of the current subject matter, the input / output device 540 can provide input / output operations for a network device. For example, the input / output device 540 can include Ethernet ports or other networking ports to communicate with one or more wired and / or wireless networks (e.g., a local area network (LAN), a wide area network (WAN), and the Internet). In some implementations of the current subject matter, the computing system 500 can be used to execute various interactive computer software applications that can be used for organization, analysis and / or storage of data in various formats.

[0100] Alternatively, the computing system 500 can be used to execute any type of software applications. The user interface can be generated and presented to a user by the computing system 500 (e.g., on a computer screen monitor, etc.).

[0101]

[0098] One or more aspects or features of the subject matter described herein can be realized in digital electronic circuitry, integrated circuitry, specially designed ASICs, field programmable gate arrays (FPGAs) computer hardware, firmware, software, and / or combinations thereof.

[0102] These various aspects or features can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device. The programmable system or computing system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client- server relationship to each other.

[0103]

[0099] These computer programs, which can also be referred to as programs, software, software applications, applications, components, or code, include machine instructions for a programmable processor, and can be implemented in a high-level procedural and / or object- oriented programming language, and / or in assembly / machine language. As used herein, the term “machine-readable medium” refers to any computer program product, apparatus and / or device, such as for example magnetic discs, optical disks, memory, and Programmable Logic Devices (PLDs), used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term “machine-readable signal” refers to any signal used to provide machine instructions and / or data to a programmable processor. The machine -readable medium can store such machine instructions non-transitorily, such as, for example, as would a non-transient solid- state memory or a magnetic hard drive or any equivalent storage medium. The machine -readable medium can alternatively or additionally store such machine instructions in a transient manner, such as for example, as would a processor cache or other random access memory associated with one or more physical processor cores.

[0104]

[0100] To provide for interaction with a user, one or more aspects or features of the subject matter described herein can be implemented on a computer having a display device, such as for example a cathode ray tube (CRT) or a liquid crystal display (LCD) or a light emitting diode (LED) monitor for displaying information to the user and a keyboard and a pointing device, such as for example a mouse or a trackball, by which the user may provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well. For example, feedback provided to the user can be any form of sensory feedback, such as, for example, visual feedback, auditory feedback, or tactile feedback; and input from the user may be received in any form, including acoustic, speech, or tactile input. Other possible input devices include touch screens or other touch-scnsitivc devices such as single or multi-point resistive or capacitive track pads, voice recognition hardware and software, optical scanners, optical pointers, digital image capture devices and associated interpretation software, and the like.

[0105]

[0101] In the descriptions above and in the claims, phrases such as “at least one of’ or “one or more of” may occur followed by a conjunctive list of elements or features. The term “and / or” may also occur in a list of two or more elements or features. Unless otherwise implicitly or explicitly contradicted by the context in which it is used, such a phrase is intended to mean any of the listed elements or features individually or any of the recited elements or features in combination with any of the other recited elements or features. For example, the phrases “at least one of A and B;” “one or more of A and B;” and “A and / or B” are each intended to mean “A alone, B alone, or A and B together.” A similar interpretation is also intended for lists including three or more items. For example, the phrases “at least one of A, B, and C;” “one or more of A, B, and C;” and “A, B, and / or C” are each intended to mean “A alone, B alone, C alone, A and B together, A and C together, B and C together, or A and B and C together.” Use of the term “based on,” above and in the claims is intended to mean, “based at least in part on,” such that an unrecited feature or element is also permissible.

[0106]

[0102] The subject matter described herein can be embodied in systems, apparatus, methods, and / or articles depending on the desired configuration. The implementations set forth in the foregoing description do not represent all implementations consistent with the subject matter described herein. Instead, they are merely some examples consistent with aspects related to the described subject matter. Although a few variations have been described in detail above, other modifications or additions are possible. In particular, further features and / or variations can be provided in addition to those set forth herein. For example, the implementations described above can be directed to various combinations and sub combinations of the disclosed features and / or combinations and sub combinations of several further features disclosed above. In addition, the logic flows depicted in the accompanying figures and / or described herein do not necessarily require the particular order shown, or sequential order, to achieve desirable results. For example, the logic flows may include different and / or additional operations than shown without departing from the scope of the present disclosure. One or more operations of the logic flows may be repeated and / or omitted without departing from the scope of the present disclosure. Other implementations may be within the scope of the following claims.

[0107] References:

[0108] [1] Haddadin, Z., Khan, S., & Poulikakos, L. V. (2023). Cutting comers to suppress high-order modes in Mie resonator arrays. ACS Photonics, 77(1), 187-195.

[0109] [2] Gomez-Robledo, L., Valero, E. M., Huertas, R., Martinez-Domingo, M. A., & Hernandez- Andres, J. (2018). Do EnChroma glasses improve color vision for colorblind subjects? Optics Express, 26(22), 28693-28703. https: / / doi.org / ] 0.1364 / oe.26.028693

[0110] [3] Roostaei, N., & Hamidi, S. M. (2022). Plasmonic eyeglasses based on gold nanoparticles for color vision deficiency management. ACS Applied Nano Materials, 5(12), 18788-18798.

[0111] [4] Tian, Y., Tang, H., Kang, T., Guo, X., Wang, J., & Zang, J. (2022). Inverse-designed aid lenses for precise correction of color vision deficiency. Nano Letters, 22(5), 2094-2102.

[0112] [5] Tsekouras, G. E., Rigos, A., Chatzistamatis, S., Tsimikas, J., Kotis, K., Caridakis, G., & Anagnostopoulos, C. N. (2021). A novel approach to image recoloring for color vision deficiency. Sensors, 21(8), 2740.

[0113] [6] Machado, G. M., Oliveira, M. M., & Fernandes, L. A. (2009). A physiologically-based model for simulation of color vision deficiency. IEEE transactions on visualization and computer graphics, 75(6), 1291-1298.

[0114] [7] Yang JH, Babicheva VE, Yu MW, Lu TC, Lin TR, Chen KP. Structural Colors Enabled by Lattice Resonance on Silicon Nitride Metasurfaces. ACS Nano. 2020 May 26; 14(5):5678— 85.

Claims

CLAIMSWhat is claimed:

1. A method of designing a filter for treating Color Vision Deficiency (CVD), the method comprising: converting cone responses for Normal Vision (NV) to the brain color coordinate system (BCCS); converting cone responses of a person with CVD to BCCS; selecting at least one coordinate of the color coordinate system; for at least one wavelength for the at least one coordinate, determine a ratio between a magnitude for the coordinate for NV and the magnitude for the corresponding coordinate for the person with CVD; and designing an attenuation profile that, when combined with the at least one CVD coordinate at the at least one wavelength, will cause the coordinate of the combination to be closer to the corresponding NV coordinate than the corresponding coordinate of the person with CVD.

2. The method of claim 1 wherein the selected coordinate is the RG coordinate of BCCS.

3. The method of claim 1 further comprising: fabricating a filter having the designed attenuation profile.

4. The method of claim 1 wherein the at least one wavelength comprises at least one wavelength band within the visible light spectrum and wherein the attenuation profile causes a plot of the coordinates as a function of wavelength in the band for the combination to be, on average, closer to a plot of the coordinates for the corresponding wavelengths for NV than a plot of the coordinates the corresponding wavelengths for the CVD person.

5. The method of claim 4 further comprising:prior to the determining step, ratiometrically scaling the NV plot down to increase the portion of the band over which a magnitude of the NV plot is less than a magnitude of the CVD plot.

7. A method of designing a filter for treating Color Vision Deficiency (CVD), the method comprising: converting cone responses for Normal Vision (NV) to the brain color coordinate system (BCCS) for a predetermined spectrum; converting cone responses of a person with CVD to BCCS for the predetermined spectrum; converting the coordinates in the brain color coordinate system for NV to CIELAB color coordinate space; converting the coordinates in the brain color coordinate system for the person with CVD to CIELAB color coordinate space; calculating the distance in CIELAB coordinate space between the coordinates for NV and the corresponding coordinates for the person with CVD over the predetermined spectrum; determining a scaling factor over the predetermined spectrum that minimizes the difference between the NV coordinates and the coordinates for the person with CVD over the predetermined spectrum.

8. The method of claim 7 further comprising: fabricating a filter effectuating the scaling factor over the predetermined spectrum.

9. The method of claim 7 wherein the predetermined spectrum is isoluminant and monochromatic.

10. The method of claim 7 wherein the predetermined spectrum has the same power at every wavelength.

11. The method of claim 7 wherein the predetermined spectrum is a natural spectrum.

12. The method of claim 7 wherein the predetermined spectrum is a spectrum based on paint chips.

13. The method of claim 7 wherein the predetermined spectrum is generated without illumination.

14. The method of claim 7 wherein the predetermined spectrum is generated with illumination.

15. The method of claim 14 wherein the illumination is standard illumination D65.

16. An optical filter for affecting color vision, the filter comprising: a pass-band between about 400 nm and about 550 nm an attenuation notch between about 550 nm and 585 nm; and a pass band between about 590 nm and 700 nm.

17. The optical filter of claim 16 wherein the attenuation band has a peak attenuation of at least 90 percent.

18. The optical filter of claim 17 wherein the peak attenuation is at about 585 nm.

19. The optical filter of claim 17 wherein the attenuation in the attenuation band increases at rate of about 2.5 percent per nm of wavelength between about 550 nm and about 585 nm.

20. The optical filter of claim 19 wherein the attenuation in the attenuation band decreases at a rate of about 90 percent per nm of wavelength between about 585 nm and about 590 nm.

Images