Biologically Inspired Enhanced Elastic Graph Matching (EEGM) Used in Object Recognition System

P.A. Martínez-Ruiz, M. Pinzolas Prado, and J. López Coronado (Spain)

Keywords

Image Features, Object Recognition, Elastic Graph Matching

Abstract

In this work, an enhanced Elastic Graph Matching algorithm is presented and tested for the recognition of several geometric figures in the workspace of a robotic platform. The purpose of this vision system is to localize and recognize not only the position, but also the orientation of the object to be manipulated. The original Elastic Graph Matching has been modified to increase the simplicity and efficiency by means of using a direct comparison of the nodes’ information instead of using more complex features that require a higher computational cost without actually increasing recognition accuracy. With the purpose of reducing the background influence, a modification is presented in the matching of the nodes. To increase the efficacy, a reinforcement method between candidates is created. The results obtained are shown in this paper.

Important Links:



Go Back


IASTED
Rotating Call For Paper Image