Developing a New Method of Transformation for Obtaining XYZ Color Values from RGB Images for Agricultural Applications

The extraction of device-independent color values from affordable and accessible digital images based on a standard color space system is crucially necessary for agricultural applications, where color information for plant monitoring or diagnostics is required. This study aimed to develop a transformation matrix for obtaining XYZ color coordinates from the RGB values of digital images for agricultural applications. The calibration procedure was based on Munsell and Macbeth color charts. The color coordinates of eight color charts were measured, and the transformation matrices were built. Leaf samples of six different plants were used and compared based on the proposed transformation technique. The actual XYZ values of plant leaves were measured, and the RGB values were derived from the digital images. The results indicate that the Macbeth color chart with 24 colors had the best performance, with an average ∆ELAB and CIEDE2000 of less than 1.77 and 1.97, respectively. The findings demonstrate that the proposed transformation matrix was successful in converting RGB values to XYZ values and can be employed as a quick, easy, and inexpensive technique for obtaining standard color information.