This paper addresses the issues of compression and computationefficiency of transform-based image compression methods. Mosttransform-based compression systems operate based on the assumption thatthe input images are mostly conventional types of images. We argue thatthese systems will not perform efficiently in some image domains inwhich the aforementioned assumption does not hold. We believe an imageencoding system must take into consideration the directional andfrequency attributes in the input images in order to achieve highencoding quality. We present a novel adaptive approach for imagetransformation. Experimental results show that for an“unconventional” image, the proposed system produces betterimage quality than the JPEG at the same bit-rates
展开▼