Optical filters affecting color vision in a desired manner and design method thereof by non-linear optimization

A computer implemented method for designing an optical filter that affects color vision in a desired manner, the method comprising iteratively evaluating a process until the luminous transmittance of the resulting optical filter defined as the weighted photopic transmittance through the lens of CIE Standard Illuminant D65 according to the CIE 1931 2-degree Standard Observer is about equal to a target luminous transmittance, the iterated process comprising:

• simulating a current filter based on a current dye formula,
• testing the filter against a chromaticity constraint criterion,
• if said test is found in violation of said constraint criterion then improving the dye formula with respect to a constraint violation measure,
• if said test is found not in violation of said constraint criterion then improving the dye formula according to a colorimetric performance measure,
• simulating a set of candidate filters based on a set of candidate incremental changes to the current dye formula,
• said improvements calculated using a ratio of the change of said measures with respect to the change in luminous transmittance wherein the changes are calculated with respect to the differences between the properties of the current filter and with respect to each candidate filter,
• updating the current filter to the candidate filter that maximizes said ratio, and, in the iterated process, the current dye formula is initialized using an initial dye formula, and the resulting optical filter is the current filter after the iterations are completed, and wherein:
• the optical filter's transmission of light as a function of wavelength, τF, is given by the expression
• τF=10(−1×F)
• F=α1Ω1+ . . . αNΩN
• and in the above expression F is the optical density spectrum of the optical filter, α1 . . . αN are dye concentrations and Ω1 . . . ΩN are optical density spectra of a set of dyes, and N is greater than 1.