Computing quadtree medial axis transform by a multi-layered pyramid of LISP-processor arrays

The authors explore the potential of using the quadtree image-representation scheme in a multiprocessor pyramid and present algorithms for computing the quadtree medial-axis transform of a binary image on the basis of a bottom-up multilayered pyramid of LISP-processor arrays. All the data structures are represented in the list form. There are three layers of processors at each level of the pyramid: PYRAMID layer, DIST layer, and QMAT layer. The first two layers form the quadtree data structure and compute the chess-board distance transform values, and the QMAT layer computes the quadtree medial-axis transform