A class of fast algorithms for the Peano-Hilbert space-filling curve

Time- and space-efficient nonrecursive algorithms, formulated directly in terms of bitwise operations, are developed to address a number of practical image scanning problems involving the Peano-Hilbert space-filling curve. These include a generator for the curve and algorithms for two-way conversion between the coordinates of a pixel and its position in the Peano-Hilbert scan. The conversion algorithms in turn lead to fast algorithms for determining the predecessor and successor of a pixel in the scan. Performance of the algorithms is assessed via analysis of their memory usage and computational requirement.<<ETX>>