Automated Phenotyping of Perennial Food Crops

Overview

This project develops automated phenotyping systems using computer vision and machine learning to accelerate breeding programs for perennial food crops, with a focus on avocado.

Funding: ~$86,000 USDA NIFA subaward (2023-2024) Role: Co-Investigator

The Problem

Traditional phenotyping of perennial crops like avocado is labor-intensive and time-consuming, limiting the scale and speed of breeding programs needed to develop climate-resilient varieties.

Our Approach

• Computer Vision: Automated image analysis for trait extraction
• High-Throughput Phenotyping: Scalable data collection pipelines
• Machine Learning Models: Trait prediction and selection optimization
• Integration: Connecting phenotypic data with breeding decision-support systems

Key Outcomes

• High-throughput phenotyping using computer vision
• Automated data collection and analysis pipelines
• Machine learning models for trait prediction
• Integration with breeding program workflows

Impact

Accelerating selection cycles and improving precision in trait evaluation for perennial crop breeding, supporting the development of climate-resilient avocado varieties.