Multiple images are combined where the images exhibit variations in illumination between one another. The images are of the same portion of an object, and each image is represented by a set of image data. A comparative data transform is applied to each set of image data, such that the transformed comparative image data isolates and preserves first variations in the illumination between the images but suppresses second differences in illumination between the images. At least one normalization transform is determined from the transformed comparative image data sets. When applied to at least one of the image data sets, the at least one normalization transform minimizes the variation in illumination between the image data sets. Each determined normalization transform is applied to the at least one of the image data sets. The normalized sets of image data are transformed to a single image of the portion of the object.