Red-and-green-based pseudo-RGB color models for the comparison of digital images acquired under different brightness levels

AbstractDigital images of crops and other plants are useful in plant management although the brightness of digital images varies with time. Brightness adjustment of the entire area of color digital images has been quite difficult due to the absence of correlations among color components and brightness. Redness and greenness best correlate with brightness. Hence, red- and green-based pseudo-color models could enable the adjustment of brightness of the entire area of color digital images. A pseudo-color model, a red–green–white mat model, lost most of the information on yellowness when tested with a color gamut. However, a red–green-inverted [red + green] model largely retained the original information. In this color model, the intensity of blue becomes the indicator of darkness. As a complementary color model, a red–green–[red + green] color model was examined. The color model provided pseudo-color images carrying information on redness, greenness, and yellowness. By comparing grayscale images of a single color component such as key black for the pseudo-color models, the pseudo-color models were shown to provide measures that are dimensionally independent of one another. Thus, together with the original red–green–blue (RGB) color image, the pseudo-color images almost triple the amount of information carried by the grayscale images. These multidimensional pieces of information are expected to facilitate the observation of colors of plants through digital image acquisition.Keywords: brightness adjustmentcrops and plantsdigital imagesred–green–blue (RGB) color modelRGB yellowstandard color