Minimum negativity constraint applied to fluorescence color signal recovery

Abstract

Color signal for a fluorescing object is the product of its emittance spectrum and the transmittance spectrum of the filter used to observe it. The color signals of microscopic fluorescing samples recovered from their image pixel color components (red, green and blue or RGB) by a transformation matrix that maps RGB to basis spectra coefficients, exhibit negative values due to the limited number of basis spectra used. A minimum negativity constraint is applied to compute the coefficients of succeeding basis spectra that will produce a reconstructed color signal having less negative values. The technique is demonstrated on fluorescent microspheres and leaf chloroplast in an actual experimental setup.