A spectral filter is designed for enhancing the visual discrimination between two colors in a dataset. The filters are designed in spectral domain using spline interpolation through maximazing the color difference between the two selected colors. Two spectral datasets were applied in the design, the first dataset contained skin and vein colors from human arms. The second dataset consisted of facial colors with and without cosmetics. In the experiments the filters for both datasets were designed and the computational results showed clear enhancement in color differences for easier discrimination between the colors. The numerical results indicated that the proposed approach is reasonable for discrimination enhancement. The visual inspection of the RGB color images showed also enhancement even though the color differences in the test images were close to just-noticiable differences.