1993Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIERequires access

Problem of determining whether a parallel reduction operator for n -dimensional binary images always preserves topology

T. Yung Kong

Open publisher page 45 citations

Abstract

Loosely speaking, a simple set of a finite binary image is a set of 1s whose deletion `preserves topology.' This concept can be made precise in different (and inequivalent) ways. Ronse established results which imply that, for finite 2-D binary images on a Cartesian grid and three different definitions of simple set, a set S of 1s is simple if every subset of S that lies in a 2- point by 2-point square is simple. In fact this is a special case of a general result which applies to arbitrary finite binary images -- not just 2-D images on a Cartesian grid -- and any definition of simple set which satisfies three axioms stated in this paper. For finite binary images on an n-dimensional Cartesian grid, we give appropriate definitions of simple set which satisfy all the axioms. When these definitions of simple set are used, verification that a parallel reduction operator for n-dimensional binary images preserves the topology of all possible input images may be achievable by checking only a finite number of cases.

About this research paper

What this paper is about

Loosely speaking, a simple set of a finite binary image is a set of 1s whose deletion `preserves topology.' This concept can be made precise in different (and inequivalent) ways. Ronse established results which imply that, for finite 2-D binary images on a Cartesian grid and three different definitions of simple set, a set S of 1s is simple if every subset of S that lies in a 2- point by 2-point square is simple. In fact this is a special case of a general result which applies to arbitrary finite binary images -- not just 2-D images on a Cartesian grid -- and any definition of simple set which satisfies three axioms stated in this paper. For finite binary images on an n-dimensional Cartesian grid, we give appropriate definitions of simple set which satisfy all the axioms. When these definitions of simple set are used, verification that a parallel reduction operator for n-dimensional binary images preserves the topology of all possible input images may be achievable by checking only a finite number of cases.

Why it matters

OpenAlex reports 45 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Loosely speaking, a simple set of a finite binary image is a set of 1s whose deletion `preserves topology.' This concept can be made precise in different (and inequivalent) ways. Ronse established results which imply that, for finite 2-D binary images on a Cartesian grid and three different definitions of simple set, a set S of 1s is simple if every subset of S that lies in a 2- point by 2-point square is simple. In fact this is a special case of a general result which applies to arbitrary finite binary images -- not just 2-D images on a Cartesian grid -- and any definition of simple set which satisfies three axioms stated in this paper. For finite binary images on an n-dimensional Cartesian grid, we give appropriate definitions of simple set which satisfy all the axioms. When these definitions of simple set are used, verification that a parallel reduction operator for n-dimensional binary images preserves the topology of all possible input images may be achievable by checking only a finite number of cases.

Key concepts: Simple (philosophy), Binary number, Regular grid, Axiom, Computer science, Finite set, Binary image, Topology (electrical circuits)

Related papers

Back to paper searchBrowse research topicsOriginal source
Problem of determining whether a parallel reduction operator for n -dimensional binary images always preserves topology — Research Paper | ScholarLens