Spectral image dissociation/reconstruction involves using image pyramids, selecting minimal transformation parameter combination, forming next pyramid image, difference images

摘要

The method involves using an image pyramid, selecting a defined minimal combination of parameters for the transformation of sub-areas given by dividing the image and the subsequent pyramid image into minimal sub-areas of 2x2 pixels each, forming the mean to form the next pyramid image from the 4 pixels (maximal decimation) and determining 3 difference value images in each pyramid plane, each of 4 pixel positions at most. An Independent claim is also included for an image processing device.