Background 

Plant phenotyping is the process of characterising visible traits in plants such as height, leaf colour, vegetation indices, and canopy surface area. It can assist researchers to understand the relationship between genes and these observable characteristics, valuable information to accelerate plant breeding programmes. However, plant phenotyping under field conditions is commonly a bottleneck in genotype–phenotype studies as it is costly, labour-intensive and prone to error. Large-scale phenotyping and trait analysis solutions can contribute to alleviate this bottleneck. Prof Ji Zhou, Head of Data Sciences Department at Cambridge Crop Research NIAB, focuses on developing large-scale phenotypic analytic solutions using computer vision and artificial intelligence techniques. Below, three of these phenotyping and analytic tools are described: SeedGerm, AirMeasurer, and CropQuant-3D. Prof Zhou’s labs in the UK and China (Nanjing Agricultural University) are always looking for partners who are passionate about sharing, or collaborators that could benefit from the research-based toolkits developed by his groups.

Target user

Researcher communities and commercial use.

Open source choice

“We are computational biologists, and open software is more secure and stable than proprietary software as we could view and modify the source code to tailor analytic solutions for trait analysis. Also, we could correct errors based on the original source code,” says Prof Zhou. Another advantage of being open source is the modular design approach, and hence researchers can use just part of the modules for their own research needs.