作者: A. Carmona-Poyato , R. Medina-Carnicer , F.J. Madrid-Cuevas , R. Muñoz-Salinas , N.L. Fernández-García
DOI: 10.1016/J.PATCOG.2010.07.029
关键词: Signal classification 、 Similarity (geometry) 、 Algorithm 、 Similitude 、 Approximation algorithm 、 Iterative method 、 Mathematical optimization 、 Mathematics
摘要: This paper presents a novel method for assessing the accuracy of unsupervised polygonal approximation algorithms. measurement relies on called ''reference approximation''. The reference is obtained using Perez and Vidal [11] by an iterative that optimizes objective function. Then, proposed calculated comparing with to be evaluated, taking into account similarity between original contour, penalizing approximations excessive number points. A comparative experiment commonly used algorithms showed more efficient than other measurements at different