OPTIMAL CLASSIFICATION USING RBF FOR FACE RECOGNITION

Classification analysis work performed by radial based function networks (RBF). I watched to obtain a minimum number of incorrect classifications based on image processing using features extraction algorithm using a variable number of pixels in each image analysis. I determined the optimal performan...

Full description

Saved in:
Bibliographic Details
Published in:Scientific Bulletin ("Mircea cel Bătrân" Naval Academy) 2015-07, Vol.18 (2), p.366-366
Main Authors: Pirloaga, Marian Dorin, Rogobete, Marius, Racuciu, Ciprian, Cretu, Emil
Format: Article
Language:English
Subjects:
Classification
Feature extraction
Image processing systems
Optimization
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Classification analysis work performed by radial based function networks (RBF). I watched to obtain a minimum number of incorrect classifications based on image processing using features extraction algorithm using a variable number of pixels in each image analysis. I determined the optimal performance for a minimum number of pixels processed and RBF unit for radius. This was achieved by two representations of data: Gaussian function, Euclidean distance and Gaussian function, Manhattan distance. At the same time I realized and a representation of performance classification by radius, number of RBF units and processing time. Finally we concluded the best efficacy experiment.
ISSN:1454-864X
2392-8956