METHOD FOR COMPRESSING IMAGES AND A FORMAT FOR COMPRESSED IMAGES

摘要

There is provided a method for compressing images and an image format. The method comprises performing a transform on pixel portions of image data and evaluating resulting coefficients from each transform in terms of a flatness condition. If the transform is found to be flat then the result from the transform is quantized according to a first quantizing scheme. If the transform is found not to be flat then the result from the transform is quantized according to a second quantizing scheme being different from the first quantizing scheme. Different block arranging schemes are proposed for the quantizing schemes. An indicator is stored in the block indicating that the image block is flat or not flat. The resulting bits are written into a compression data structure.