Texture control in image quantization by iterative wavelet transform algorithms

The simplicity with which information can be extracted from various data depends on the representation of the data. The information contents of images can be represented in various ways, e.g., as the image itself or as the Fourier spectrum of the image. Manipulations can be applied in any of these representations. Among the algorithms for image quantization, the image and the spectrum representations are widely utilized. An iterative quantization algorithm is presented that uses the wavelet representation of the image. This algorithm allows control of the local texture of the quantized image. © 1996 Optical Society of America.