Classifying master images method for electronic reproduction, according to gray- and color-images, involves providing image points as image point vectors in a color system with several color components

摘要

The method of classifying master images, illustrations etc is a development of the earlier used method based simply on the examining the difference between gray images and color images, where the red-green-blue (RGB) components were examined in terms of whether most of the image points of the RGB-values were approximately the same size or varied greatly. The image points are now given as image point vectors (rx) in a color system with several color components according to gray images and color images. The covariance matrix C is determined from the image point vectors (rx), and the intrinsic values (lambda i ) and the intrinsic vectors (ei) of the covariance matrix C are found for classifying the master images on the basis of lambda i and C.