DEEP LEARNING ALGORITHMS FOR AUTOMATIC DETECTION AND CLASSIFICATION OF MILDEW DISEASE IN CUCUMBER

Ozguven M. M.

FRESENIUS ENVIRONMENTAL BULLETIN, cilt.29, sa.8, ss.7081-7087, 2020 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 29 Sayı: 8
  • Basım Tarihi: 2020
  • Dergi Adı: FRESENIUS ENVIRONMENTAL BULLETIN
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Agricultural & Environmental Science Database, Aqualine, Aquatic Science & Fisheries Abstracts (ASFA), CAB Abstracts, Chemical Abstracts Core, Communication Abstracts, Environment Index, Geobase, Greenfile, Metadex, Pollution Abstracts, Veterinary Science Database, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.7081-7087
  • Anahtar Kelimeler: CNN, cucumber, Faster R-CNN, mildew disease, Pseudoperonospora cubensis
  • Ankara Üniversitesi Adresli: Hayır

Özet

This study aims to contribute literature on the detection of mildew disease, determination of the disease severity, and real-time and automatic monitoring of the course of the disease using cucumber leaf images. For this purpose, Faster R-CNN model was used, which is an object detection and localization method based on the Convolutional Neural Network architecture, a state-of-the-art deep learning method, The proposed method was trained and tested with a total of 175 images. As a result of tests performed to determine disease and disease severity in cucumber leaves, a correct classification rate of 94.86% was achieved in total. The proposed method was found to be successful in terms of these metrics, compared to contemporary methods in the literature. In addition, sample collection and laboratory analysis become unnecessary since the proposed method has a practical application that allows automatic acquisition and evaluation of the images of the plants via imaging devices of various resolutions. Therefore, the proposed method is believed to be beneficial in rapid detection of other plant diseases and pests, and to determine disease severity and to monitor the progress of the disease, as the study advances, despite the current focus on the cucumber leaves.