Enhanced Crop Scanning System (CSS) for Optimized Cucurbit Production

Semi-weekly fungicide applications throughout the season are essential to ensure profitable cucurbit yields. Lack of adequate disease control decreases fruit number, fruit weight, and fruit quality. In the proposed Phase I work we will first focus on watermelon to demonstrate feasibility for other cucurbit crops. There is evidence that information collected by scouting watermelons for disease symptoms increases the efficacy of fungicide sprays by refining the choices of fungicide products and timings. Manual scouting is expensive, slow, and incomplete, typically exploring less than 1% of the crop area and resulting in a low probability of early disease detection. The Crop Scanning System (CSS) proposed herein will digitally scout every watermelon plant in every field during routine weekly fungicide spraying operations. The specific objectives are: 1) Design a tractor-mounted CSS utilizing machine-vision cameras for collection and analysis of georeferenced images in watermelon fields, 2) Develop an automated Cloud-based repository and Web-GIS app to map watermelon vines, fruit, disease symptoms, weeds and anomalies with the CSS, 3) Evaluate the CSS camera system and Cloud GIS as a digital scouting tool for enhanced detection and mapping of early-stage foliar disease and anomalies on watermelon. This will be the first ground-based digital scouting tool developed for cucurbits to improve disease control, marketable yields and reduce labor costs.