Vol.26, No.4, 2020, pp.693-702, doi:10.32604/iasc.2020.010103
OPEN ACCESS
ARTICLE
Object Detection and Fuzzy-Based Classification Using UAV Data
  • Abdul Qayyum1,*, Iftikhar Ahmad2, Mohsin Iftikhar3, Moona Mazher4
1 mViA laboratory, University of Bourgogne Franche-Comté, 21000, Dijon, France
2 Department of Software Engineering, College of Computer and Information Sciences, P.O. Box 51178, Riyadh 11543, King Saud University, Riyadh, KSA
3 Centre School of Computing and mathematics, Charles Sturt University, Australia
4 Centre for Intelligent Signal and Imaging Research (CISIR), Department of Electrical and Electronic Engineering, Universiti Teknologi PETRONAS, 32610 Bandar Seri Iskandar, Perak, Malaysia.
* Corresponding Author: Abdul Qayyum, engr.qayyum@gmail.com
Abstract
UAV (Unmanned Aerial Vehicle) equipped with remote sensing devices can acquire spatial data with a relevant area of interest. In this paper, we have acquired UAV data for high voltage power poles, urban areas and vegetation/trees near power lines. For object classification, the proposed approach based on the fuzzy classifier is compared with the traditional minimum distance classifier and maximum likelihood classifier on our three defined segments of UAV images. The performance evaluation of all the classifiers was based on the statistics parameters which included the mean, standard deviation and PDF (probability density function) of each object present in the image acquired by the UAV and the variances of each channel of the UAV imagery were calculated. The results showed that the fuzzy-based classifier outperformed as compared to the other classifiers. We achieved the classification accuracy of 93% with a Fuzzy-based classifier.
Keywords
Classifier, Fuzzy Logic, Minimum Distance, Orthorectification, Spectral property, Supervised Classifier, UAV.
Cite This Article
A. Qayyum, I. Ahmad, M. Iftikhar and M. Mazher, "Object detection and fuzzy-based classification using uav data," Intelligent Automation & Soft Computing, vol. 26, no.4, pp. 693–702, 2020.
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.