ManziniTech1110/NAILit-Maize-Disease-Analyser

Graph-based maize leaf disease detection for South African farmers # ⚡ NAILit — Maize Leaf Disease Analyser > **v2.0 | Graph-Based Disease Detection System** > A JavaFX desktop application for maize leaf disease classification, similarity analysis, and leaf health reporting. --- ## 📋 Table of Contents - Overview - Features - System Architecture - Disease Classification - Tabs & Panels - How It Works - Project Structure - Requirements - Running the Application - Team --- ## Overview **NAILit** is a desktop tool built for agricultural researchers and farmers to detect and analyse diseases in maize (*Zea mays*) leaf images. It uses graph-based image segmentation, BFS cluster analysis, and seasonal colour-aware classification to deliver: - Automated disease classification per uploaded image - Per-disease confidence intervals on RGB colour channels - Pairwise image similarity scoring - Season-aware detection across Summer, Autumn, Spring, and Winter --- ## Features | Feature | Description | |---|---| | 🔬 **Disease Classification** | Classifies images as Healthy, Northern Leaf Blight, Common Rust, or Gray Leaf Spot | | 📊 **Similarity Matrix** | Pairwise vertex-to-vertex comparison across all uploaded images | | 📈 **Leaf Analyser** | BFS cluster analysis with 95% confidence intervals per disease class | | 🌱 **Season-Aware Detection** | Adjustable seasonal mode (Summer / Autumn / Spring / Winter) | | 🖼️ **Image Preview Strip** | Scrollable thumbnail bar with per-image status labels | | ⚡ **Multithreaded Analysis** | Image loading and similarity computation run off the UI thread | --- ## System Architecture ``` Main.java (UI Package — entry point) │ ├── UI Package │ ├── ClassifyPanel — Disease Classification tab │ ├── SimilarityPanel — Similarity Matrix tab │ ├── AnalyserPanel — Analyser tab (CI table + recommendations) │ ├── UIHelper — Shared static UI utilities │ └── TutorialHelper — Anime-style interactive tutorial │ ├── Graphs Package │ ├── ImageLoader — File → PixelNode …