ABSTRACT
Image encoding techniques are broadly classified under two categories, namely, first-generation image encoding techniques and second-generation image encoding techniques. The most important first-generation techniques use an orthonormal basis for representing an image, while second-generation techniques exploit image characteristics and the psychophysics of human visual perception. This chapter discusses image encoding techniques and introduce the Laplacian pyramidal algorithm of Burt and Adelson, which will be compared with the wavelet pyramidal algorithm of Mallat. The two-dimensional fast wavelet transform of a discretely sampled image is computed using the same scheme as used in the one-dimensional case. Edges in images are characterized by sharp variations in intensity values. However, these variations can occur at several scales, ranging from edges of large objects and contours of smaller objects to texture. In practice, images are obtained by discrete sampling and, therefore, have a finite resolution.
