Abstract
Downy mildew, caused by Plasmopara viticola, presents a significant challenge for vineyard protection. Conventional methods of disease detection are frequently arduous, subjective and time-consuming, which constrains their efficacy in large-scale agricultural applications. Consequently, there is an increasing requirement for technological solutions that employ artificial intelligence to improve the precision and efficacy of disease localisation and monitoring. The objective of this study was to develop an automated and interpretable approach for the detection and localisation of downy mildew symptoms in field conditions. The findings of this study demonstrate that the proposed automated methodology not only facilitates the detection of downy mildew but also provides a robust framework for accurately localising affected regions within canopy images.

The speaker
Dr. Javier Tardaguila is Full Professor of Precision Viticulture at the University of La Rioja (Spain). 
Prof. Tardaguila works in precision viticulture, robotics and emerging technologies in grape growing. He was the Coordinator of VineRobot European project and currently he is the PI of NoPest H2020 project that explores new solutions to counteract oomycete infections in agriculture. He is involved in several European and national research projects in viticulture and new technologies. He also is the author of over 85 publications in indexed journals (h index =27). Prof. Tardaguila has been actively involved in the technological transfer to the wine industry in collaboration to several national and international companies. He is Associate Editor of “Australian Journal of Grape and Wine Research”. Prof. Tardaguila has been Vice-president of Research and Knowledge Transfer at the University of La Rioja.