Goals

The goal of the project is to develop image-based machine learning methods for accurately quantifying disease severity (Fusarium wilt, nematodes, and Sigatoka) and key agronomic traits (bunch characteristics and pulp color) in bananas, enhancing phenotyping efficiency and selection accuracy in breeding programs.

Expected results

1. Identification of key traits and diseases prioritized for phenotyping using computer vision.
2. Deployment of field technicians for full-time engagement and collaboration with senior scientists/breeders.
3. Collection and sharing of image data and metadata according to quality and quantity standards.
4. Expert support provided for annotation of images.
5. Generation of a 2-page report summarizing critical learnings on SOPs, image collection, annotation, and recommendations for further work.