CLASSIFICATION OF HEALTHY AND DISEASED VINE LEAVES USING THE FULL SPECTRA OF OBJECT AREA IN IMAGE

Authors

  • K. Georgieva Faculty of Technics and Technologies, Trakia University, Bulgaria, 38 Graf Ignatiev str., 8602 Yambol, Bulgaria
  • N. Georgieva Faculty of Technics and Technologies, Trakia University, Bulgaria, 38 Graf Ignatiev str., 8602 Yambol, Bulgaria
  • Z. D. Zlatev Trakia University, faculty of technics and technologies, Yambol, Bulgaria http://orcid.org/0000-0003-3080-5048
  • G. Georgiev Faculty of Technics and Technologies, Trakia University, Bulgaria, 38 Graf Ignatiev str., 8602 Yambol, Bulgaria
  • A. Dimitrova Faculty of Technics and Technologies, Trakia University, Bulgaria, 38 Graf Ignatiev str., 8602 Yambol, Bulgaria

DOI:

https://doi.org/10.4314/jfas.v10i3.3

Keywords:

grape leaves, full spectra of image, spectral range, discriminant function analysis

Abstract

Grape plant diseases cause critical harm and financial loses in crops. In this manner, early identification of diseases is important on the contemporary stage of development of science and technologies. Optical methods have been widely used to solve the task of detecting diseases in vineyards. The determination of diseases on vines by outer indications of the leaves is made by video cameras, the estimations of the colour components of various colour models are used. The disadvantage of direct using of colour components is that there are high classification errors because of complexity of colours on the surface of vine leaves. The use of full spectra of image object areas with healthy and diseased part of leaves is proposed. Lower values of classification errors are comparable with those, obtained by neural network classifier.

Downloads

Published

2018-09-01

Issue

Vol. 10 No. 3 (2018)

Section

Research Articles

License

All materials contained on this site are copyright protected and may not be reproduced; therefore, individuals may view, download, print, or save Journal content for the purposes of research, teaching, and/or private study, the Publisher holds the copyright on all materials published in website of the Journal of Fundamental and Applied Sciences . The following guidelines apply to individual users:

  1. Individuals may view, download, print, or save Journal content for the purposes of research, teaching, and/or private study.
  2. Customer or member account numbers are provided to individuals for their personal use only. Individuals may not share their username or password without the Publisher's permission.
  3. Any use and/or copies of this Journal in whole or in part, must include the customary bibliographic citation, including author attribution, date, article title, the Journal name and its web site address (http://www.jfas.info).

Disclaimer of Warranty

The Journal Online is under development and changes may be made in these publications and programs at any time.Â