Machine vision color calibration in assessing corn kernel damage

The effectiveness of many color image analysis and classification methodologies depends on the constancy of the scene illumination. In reality scene illumination often changes over time. In this article, we present a simple, bur effective method for calibration to improve color image classification with changing illumination. Color changes due to changing illumination were captured using a gray reference plate. The extent of the color changes in each of the RGB channels was calculated based on art equation derived from the spectral reflectance model. These values formed a transformation matrix which was used to transform the image RGB values to compensate for the color changes. A set of experiments for classifying the color of mold-damaged corn kernels was performed to demonstrate the effectiveness of the color calibration method in the presence of changes in illumination.