Estimation of f-Similarity in f-Triangles Using FIS

Abstract

Today, some high profile crimes grab our attention and headlines of the world. But, the core problem underlies is to identify the criminals. However, we acquire the features of miscreants narrated by spectators, the fuzzy patterns of finger prints, shoe prints and sometimes the handwriting found, are the crucial clues to apprehend the criminals. Identifying similarity of fuzzy information with the criminal database is not an easy task; this is what is being investigated in our work. We begin our work with a novel approach of estimating fuzzy similarity call it as f-similarity in fuzzy triangles using the membership values. Undoubtedly, the degrees of similarities persist in figures and hand drawn sketches, but, estimating them is performed here. In this sequel, we have discussed about f-geometry, which are the basics of f-similarity defined in terms of membership values. The membership values generated using three popularly known postulates of similar triangles like AAA, SSS and SAS are applied as inputs to the Fuzzy Inference System (FIS). We have found good results of FIS, which can be applied for any inexact geometric shape evaluation.