In the paper two LUTs are mentioned. In the code, only one is used, and one polynomial.
- where do the numbers for the weights and other coefficients come from?
- Do you guys also share the data used for the reflectance, scattering and absorption values of the all the 38 wavelength used?
- In the float_rgb_to_latent(rgb) function, the _lut seems to be a large 1D vector, as opposed to the PNG file that is a 4096x4096 2D array. Based on the paper, the _lut contains the "unmix" calculations for all RGB colors, which means the c=[c0,c1,c2] is saved in every pixel of the lut png file?
In the paper two LUTs are mentioned. In the code, only one is used, and one polynomial.