Tesseral spatio-temporal reasoning for multi-dimensional data

A generally applicable approach to N-dimensional spatial reasoning is described. The approach is founded on a unique representation based on ideas concerning "tesseral" addressing. This offers many computational advantages including minimal data storage, computationally efficient translation of data, and simple data comparison, regardless of the number of dimensions under consideration. The representation allows spatial attributes associated with objects to be expressed simply and concisely in terms of sets of addresses which can then be related using standard set operations expressed as constraints. The approach has been incorporated into a spatial reasoning system - the SPARTA (SPAtial Reasoning using Tesseral Addressing) system - which has been successfully used in conjunction with a significant number of spatial application domains. (C) 1999 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.