Skip to main content
custom logo
Improved yield estimation for the Australian wine industry
Yield estimation for wine-grapes is a challenging problem as there can be large variability from year to year in yield components, weather conditions and management actions. This webinar will present the results from a three-year Wine Australia funded study into block-level yield estimation, whereby several image processing based methods have been designed, implemented and evaluated in Australian vineyard blocks. A system of shoot counting in combination with mobile phone images of inflorescences and bunches was shown to outperform best-practice manual yield estimation at all points during the season. This opens the possibility of using simple tools across industry to reduce the tedium of performing forecasts, improve forecasting accuracy and implement management actions to optimise yield.

Nov 23, 2017 11:30 AM in Adelaide

* Required information


Dr Mark Whitty
Lecturer @University of New South Wales
Dr Mark Whitty is a Lecturer at UNSW Sydney and leads the Smart Robotic Viticulture research group. Beginning from work in autonomous robotics, he and his group have developed methods for image processing in viticulture, including shoot detection, flower counting, berry counting, yield estimation and apps for wine water stress detection.