Paper Details: Downloads: 564
Serial Number: P1121046354
Title: Sketch Recognition using Enhanced Particle Swarm Optimization
Authors: Maha El Meseery and Mahmoud Fakhr El Din and Tarek abouldahab
Abstract: Sketches and drawings are widely used as a simple method of expressing thought and designs. The goal of this paper is to improve our sketch recognition system by using an enhanced particle swarm optimization algorithm. The Enhanced Particle Swarm Optimization (EPSO) is based on the perturbation of the particle velocity is used to optimally segment the strokes the user draws into meaningful geometric primitives. These geometric primitives are grouped to formulate symbols which are further identified using a Support Vector Machines (SVM) classifier. Experiments were conducted on the benchmark dataset of simple presentation (Hs-DS)symbols. Results show that the enhanced particle swarm optimization improves the final recognition accuracy.
Keywords: Sketch Recognition, Particle Swarm Optimization
Journal/Conference: International Journal of Artificial Intelligence and Machine Learning
Volume: 11
Issue: 1
Submission Date: 11/7/2010 12:00:00 AM
Review Date: 3/18/2011 12:00:00 AM
Publishing Date: 4/17/2011 12:00:00 AM
Article Downloads: 564
Download:

Facebook