Image Coding using Fractal Parameters of Contour Lines

Abstract

Pyramidal region segmentation can be used to classify the regions of an image into two categories : the regions “not having to preserve the waveform of image” are encoded only with fractal parameters, and the regions “having to preserve the waveform of image” are encoded using any waveform coding method [DCT, VQ, etc.]. And those blocks containing region contour lines can be classified into two categories according to the complexity of the shape of contour lines : the blocks with simple contours can be encoded using the same method used for encoding the regions “not having to preserve the waveform of image”, and the blocks with complex contours can be encoded using waveform and fractal coding methods. Our experiments this combination of coding methods made it possible to encode an image at 0.4-0.7 bits/ pixel.