Fungal Pathogen Quantification Using Image Analysis in Controlled Conditions Trials

Abstract :
In recent decades, the crop protection sector has entered a period of significant change. Since the 1990s, an increasing number of plant protection products have been progressively banned, and many of the remaining compounds are now classified as candidates for substitution due to inadequate eco toxicological profiles. Simultaneously, climate change and globalization are accelerating the spreading of pathogens in new areas, while resistance mechanisms continue to develop within endogenous pathogen populations.
For several crops, this convergence of factors has led to a concerning lack of viable chemical disease management solutions. To address these challenges, lots of new strategies for pathogen management are actively explored. However, the timeline for the development of innovative solutions remains a critical constraint. Especially for perennial crops, where long growth cycles considerably slow experimentation. An efficient alternative is to conduct experiments under controlled conditions using growth chambers.
It is in this context that a method for numerical analysis of fungal disease symptoms has been developed. The proposed approach employs an RGB macro camera to acquire high-resolution images of infected leaves. Images are then analyzed through a series of colour-based filters that accurately quantify the percentage of leaf area covered by lesions. For apple scab symptom detection, the method demonstrates an error margin below 0.3%. Comparative analyses show that digital assessment is substantially more repeatable than human visual scoring and offers a more sensitive detection threshold. These advantages make the numerical evaluation method highly valuable for biofungicide development and readily applicable to other integrated pest management innovations, including varietal selection programs. Furthermore, due to its adaptable framework, the methodology can be transferred to additional pathosystems.