AI Tool Automates Plant Fruit Measurement to Improve Crop Breeding
Aberystwyth University
August 26, 2025
Scientists at Aberystwyth University have developed an artificial intelligence tool that automatically measures plant seeds and seed pods to support the breeding of improved crop varieties. This collaboration between the Institute of Biological, Environmental and Rural Sciences and the Computer Science department highlights how targeted AI applications can enhance crop quality. Their findings are published in GigaScience.
Traditional methods for recording fruit traits like shape and size are labor-intensive, time-consuming, and prone to error. The new AI-powered tool analyzes images to identify seed pods and accurately measure characteristics such as length, width, area, and volume. These traits directly influence crop yield and profitability.
The research links these physical traits with specific genetic regions that control pod shape and size. Pinpointing these genes helps scientists understand plant development and identify targets for breeding crops with improved yield, form, and resilience.
This AI tool is versatile and can be applied to various crops. Researchers have tested it on oilseed rape, cabbages, and cereals including oats, barley, and wheat. Kieran Atkins, Ph.D. researcher and project lead, noted, "Our algorithm collected data on over 300,000 individual fruits, demonstrating deep learning’s ability to phenotype large populations efficiently."
Atkins added, "By eliminating technical and time barriers, this tool makes large-scale phenotyping accessible to more researchers, enabling exploration of plant traits at unprecedented scales."
Professor John Doonan, director of the National Plant Phenomics Centre, emphasized that the AI provides data quality suitable for genetic analysis and breeding. Initially developed for a model weedy plant, similar methods work effectively on brassica crops. This marks a step toward scalable, data-rich phenotyping that accelerates research and supports predictive crop improvement.
The MorphPod tool is publicly available online, allowing researchers worldwide to replicate or adapt it for other plant species. For more details, see the published study by Kieran Atkins et al., Unlocking the power of AI for phenotyping fruit morphology in Arabidopsis, GigaScience (2024), DOI: 10.1093/gigascience/giae123.
Your membership also unlocks:
