Open AccessThe Hybridization of Neural Network and Particle Swarm Optimization for Natural Terrain Feature Extraction
Author(s) -
Sakshi Dhingra,
P. Suresh Kumar
Publication year - 2019
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.a4822.119119
Subject(s) - particle swarm optimization , metaheuristic , terrain , computer science , artificial neural network , cohen's kappa , multi swarm optimization , swarm behaviour , data mining , artificial intelligence , feature (linguistics) , pattern recognition (psychology) , geospatial analysis , soft computing , remote sensing , algorithm , machine learning , geography , linguistics , philosophy , cartography
The optimization of various soft computing and metaheuristic techniques can be ameliorated in a global area network, Swarm intelligence. In this research, a hybrid algorithm of neural network and particle swarm optimization has been presented for remote sensing applications. The terrain features of the land in a remote sensing image have been classified using these algorithms. Remote sensing basically deals with the processing and interpretation of satellite images without any physical contact to that particular region. In addition, the geospatial characteristics of the data also recorded during image classification. The hybrid concept used in this research, the implementation of algorithm in this paper based on the neurons network to find the best solution, which is further resolved using the Particle Swarm Optimization approach, an optimization technique. The proposed algorithm easily classifies the terrain features with higher efficiency and kappa coefficient value. The results show that 94.36% accuracy attained from the proposed technique. The overall accuracy improved by 5.24 % and 14.93% and kappa coefficient enhancement of 6.97 % and 18.99 % in comparison to existing studies.