The Utility of RGB Color for Discrimination of Lunar Maturity and Composition

Abstract We explore the extent to which red‐green‐blue (RGB) color images, such as those produced by a complementary metal oxide semiconductor (CMOS) Bayer‐filter camera, can be utilized for studying the maturity and composition of the lunar surface. RGB filters typically are quite broad, with considerable overlap among the three colors. Convolution of laboratory spectra for lunar samples to the RGB responsivities of the Chang'E‐3 rover's Panoramic Camera allowed determination of the correlations between color ratios (B/R, B/G, and G/R) and the sample maturity (Is/FeO) or composition (wt.% FeO or TiO2). In general, color ratios decrease as Is/FeO increases. When separate sample categories are considered, we find that the B/R ratio is a good predictor of Is/FeO for low‐Ti mare, low‐Fe highland, and moderate‐Fe highland soils. For high‐Ti mare soils, Is/FeO has little influence on the B/R ratio (due to the spectral effects of abundant ilmenite), and hence the ratio cannot be used to determine maturity. We also find that color ratios have no useful correlation with sample wt.% FeO or TiO2. Thus, in locations excluding the high‐Ti maria, RGB color data could be used to estimate soil maturity. We outline a concept for a multispectral imager based on a CMOS sensor with a Bayer‐like pattern of custom‐wavelength filters chosen specifically for lunar science applications.