The Heeger & Bergen Pyramid Based Texture Synthesis Algorithm

This contribution deals with the Heeger-Bergen pyramid-based texture analysis/synthesis algorithm. It brings a detailed explanation of the original algorithm tested on many characteristic examples. Our analysis reproduces the original results, but also brings a minor improvement concerning non-periodic textures. Inspired by visual perception theories, Heeger and Bergen proposed to characterize a texture by its first-order statistics of both its color and its responses to multiscale and multi-orientation filters, namely the steerable pyramid. The Heeger-Bergen algorithm consists in the following procedure: starting from a white noise image, histogram matchings are performed to the image alternately in the image domain and the steerable pyramid domain, so that the corresponding output histograms match the ones of the input texture. Source Code An on-line demo1 of the Heeger-Bergen pyramid-based texture synthesis algorithm is available. The demo permits to upload a color image to extract a subimage and to run the texture synthesis algorithm on this subimage. The algorithm available in the demo is a slightly improved version treating non-periodic textures by a “periodic+smooth” decomposition [13]. The algorithm works with color textures and is able to synthesize textures with larger size than the input image. The original version of the Heeger-Bergen algorithm (where the boundaries are handled by mirror symmetrization) is optional in the source code. An ANSI C implementation is available for download here2. It is provided with: • An illustrated html documentation; • Source code; This code requires libpng, libfftw3, openmp, and getopt. Compilation and usage instructions are included in the README.txt file of the zip archive. The illustrated HTML documentation can be reproduced from the source code by using doxygen (see the README.txt file of the zip archive for details).