Noninferior Bandwidth Line Simplification: Algorithm and Structural Analysis

Abstract

Digital line simplification operators are an important preprocessing component of a spatial data handling system. Simplification operators are used to generalize digital line files because these files are encoded at one scale but the objects they represent can be displayed at many other scales. Most line simplification algorithms are heuristics that caricaturize lines by eliminating points representing redundant data or conversely retaining the characteristic points of a line that have a high informational content. This paper presents a noninferior simplification algorithm that produces an optimal caricature of an encoded line with respect to the bandwidth criterion that all deleted points must be within a prespecified distance of the simplified line. Results show that it is a practical alternative to existing bandwidth heuristics.