Tashkoski, Martin and Krsteska, Vesna Automatic identification of Trialeurodes vaporariorum and Bemisia tabaci. Тутун-Tobacco, 69 (1-6). pp. 40-50. ISSN 0494 - 3244
Full text not available from this repository.Abstract
The computers era and the digital revolution has made positive changes in modern life, including entertainment,
communication and science. Digital images are just one example of this revolution. Smartphones allow us to take
pictures and to modify them applying different filter. Digital image processing is used in digital cameras, to improve image quality, applying various filters that display the image in different ways, extracting information from
medical or microscopic images. Image processing often pairs with machine learning and together are powerful
method for images classification. New algorithms for image processing, their analysis and obtaining specific information are constantly being developed.
In this paper we present a method for microscopic images classification of two larvae of insects belonging to the
family Aleyrodidae, subfamily Aleyrodoidea or known as whiteflies. Both whiteflies (lat. Bemisia tabaci and
Trialeurodes vaporariorum) feed on the juices of the plant and can easily adapt on different plants. They are very
similar and can be distinguished only in a certain stage of their development (pupae stage). This paper presents the
process of taking the images, their preprocessing and processing, the creation of a descriptor that best describes the
two insect larvae, the method for cleaning noise and the background of the images, and the results obtained from
several different classification tests in Weka and SVM light.
Key words: tobacco, whiteflies, identification, distinction, processing, images
| Item Type: | Article |
|---|---|
| Subjects: | Scientific Fields (Frascati) > Agricultural Sciences > Agricultural biotechnology |
| Divisions: | Scientific Tobacco Institute |
| Depositing User: | Prof d-r. Vesna Krsteska |
| Date Deposited: | 26 Dec 2023 11:18 |
| Last Modified: | 26 Dec 2023 11:18 |
| URI: | https://eprints.uklo.edu.mk/id/eprint/8722 |
Actions (login required)
 |
View Item |