A Python-based image analysis pipeline using Mask R-CNN to extract morphological traits from grape leaf images.
This repository contains a specialized pipeline for automated analysis of grape leaf morphology. The system processes laboratory-captured leaf images to extract quantitative data on key traits including area, dimensions, lobing, and surface characteristics.
- Implement Mask R-CNN for automated extraction of leaf morphological traits
- Process historical leaf photographs from the Geneva location
- Analyze over 100 leaf samples with centimeter-per-pixel precision
- Create a reproducible workflow for future leaf analysis projects
The pipeline extracts the following leaf traits:
- Leaf area
- Length and width
- Height
- Serration patterns (leaf teeth)
- Lobe count
- Perimeter
- Venation patterns (abaxial/bottom surface)
- Color properties
- Processing of existing leaf photographs
- Python scripts implementing Mask R-CNN
- Comprehensive trait extraction algorithms
- Documentation and training materials
- Collection of new samples
- Web or mobile interfaces
- Real-time analysis capabilities
- Complete Python pipeline with Mask R-CNN implementation
- Trait-specific extraction scripts
- Configuration files for reproducible analysis
- Technical architecture documentation
- User guide and setup instructions
- Best practices for environment management
- Quantitative trait datasets
- Visualization tools
- Statistical summaries
[Installation and setup instructions will be added here]
- Python 3.10+
- Required packages listed in requirements.txt
- Command-line familiarity
Pipeline development in progress. This project builds upon existing work and data provided by Silvas Kirubakaran- Abiotic Stress Genetics Laboratory (USDA) and is coordinated by Arlyn John Ackerman. Active code development is performed by Arlyn Ackerman and Meseret Wondifraw.
[License information will be added here]