Identificator: A web-based tool for visual plant disease identification, a proof of concept with a case study on strawberry

  • Authors:
  • Ilaria Pertot;Tsvi Kuflik;Igor Gordon;Stanley Freeman;Yigal Elad

  • Affiliations:
  • Research and Innovation Centre, Fondazione Edmund Mach, via Mach 1, S. Michele all'Adige, 38010 TN, Italy;Information Systems Department, The University of Haifa, Mount Carmel, Haifa 31905, Israel;Information Systems Department, The University of Haifa, Mount Carmel, Haifa 31905, Israel;Department of Plant Pathology and Weed Research, The Volcani Center, ARO, Bet Dagan 50250, Israel;Department of Plant Pathology and Weed Research, The Volcani Center, ARO, Bet Dagan 50250, Israel

  • Venue:
  • Computers and Electronics in Agriculture
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Identificator is a web-based tool used to help non experts in identifying plant diseases, based on the selection of pictures and/or short text descriptions (when no suitable images exist) representing the symptoms on a specific sample of plant organs. The system is based on a multi-access key of identification and specifically on the selection of pictures by the user and can be used remotely from a desktop as well as from a smart phone or personal digital assistant. The system was developed following a simple approach: visual identification where images and/or short descriptions are used to uniquely identify diseases when possible and suggest refining the visual identification process in cases of ambiguous identification. It has been designed in a way that allows easy definition of additional diseases by uploading the correct images and defining the identification rules and diseases. In this way the system may aid growers in identifying various diseases when using the system remotely while the system is developed and maintained centrally. This approach may ease the process of manual visual disease identification until machine vision technology is mature enough to perform this task automatically. We tested the system for visual identification of strawberry diseases using a computer and samples of infected plants. The evaluation showed that it is effective and accurate in enabling its users to identify strawberry diseases.