## Abstract

Parallel algorithms for quadtree problems are developed for a computing structure with a single, fixed communication topology--two-dimensional shuffle-exchange. Quadtree problems are defined as those that can be solved by using quadtree as the data structure. The key characteristic of a quadtree structure is its recursive subdivisions of the image into quadrants. This subdivision pattern can be provided naturally by the two-dimensional shuffle-exchange network (2-D SE). It is shown that algorithms for quadtree problems, such as area computation, centroid, and perimeter calculation, can be solved with the time complexity of O(log N) if the size of the image is N multiplied by N. Various implementations for 2-D SE are presented.

Original language | English |
---|---|

Title of host publication | Unknown Host Publication Title |

Publisher | IEEE |

Pages | 140-147 |

Number of pages | 8 |

ISBN (Print) | 0818607211 |

Publication status | Published - 1986 |

## All Science Journal Classification (ASJC) codes

- General Engineering