Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
Vision-based Grapevine Pierce’s disease detection system using artificial intelligence
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: 2018 ASABE Annual International Meeting 1800148.(doi:10.13031/aim.201800148)
Authors: Albert C Cruz, Ashraf El-Kereamy, Yiannis Ampatzidis
Keywords: Pierce’s Disease, Vitus vinifera L., deep learning, artificial intelligence, machine learning
Abstract. Xylella fastidiosa is a bacterial pathogen responsible for Pierce‘s Disease (PD) in Vitus vinifera L. Infections of PD remain a problem, despite efforts by the CFDA and USDA to control the disease vector. In 2014, PD caused $104.4 million losses to Californian wine and table grape growers. This work details a system to detect PD automatically by computers. The proposed system can provide early screening to improve reaction time of control strategies because it can be deployed earlier than conventional lab tests and at a lower cost. Due to xylem localization, petioles are collected for testing after leaf blades have dried. This causes delays and further contamination of neighboring vines. However, PD infected vines display leaf scorching which can be automatically identified from leaf clippings three to eighteen months after initial infection. Preliminary results with a prototype deep learning system (AlexNet) have a sensitivity of 98.99%. We are the first to automate grapevine PD detection with deep learning algorithms. When implemented on hand-held diagnostic tools, the technologies pioneered in this program can accelerate a response to mitigate crop losses of other diseases such as phony peach disease, citrus variegated chlorosis, and others.
(Download PDF) (Export to EndNotes)