Advances in digital camera-based phenotyping of Botrytis disease development

Abstract

Botrytis cinerea is an important generalist fungal plant pathogen that causes great economic losses. Conventional detection methods to identify B. cinerea infections rely on visual assessments, which are error prone, subjective, labor intensive, hard to quantify, and unsuitable for artificial intelligence (AI) and machine learning (ML) applications. New, often camera-based, techniques provide objective digital data by remote and proximal sensing. We detail the B. cinerea infection process and link this with conventional and novel detection methods. We evaluate the effectiveness of current digital phenotyping methods to detect, quantify, and classify disease symptoms for disease management and breeding for resistance. Finally, we discuss the needs, prospects, and challenges of digital camera-based phenotyping.