Sahil-web01/AgriEdge

# 🌿 AgriEdge: Offline-First Crop Health Intelligence ### Edge AI Diagnostics, Disease Severity Grading & Real-Time NGO Command Center **Zero-Cloud Edge Crop Diagnostics for Smallholder Farmers** -059669?style=for-the-badge&logo=speedtest&logoColor=white) -8B5CF6?style=for-the-badge&logo=soundcharts&logoColor=white) --- ## 📌 Hackathon Project Overview - **Event:** OMNIKON National Hackathon 2026 (Software Track) - **Theme:** AgriTech & FoodTech - **Problem Statement Code:** `Omni_AgriTech_2: Offline-Capable Crop Disease Detection` - **Core Challenge:** Smallholder farmers in connectivity-deprived rural zones lack real-time diagnostic support when crop disease strikes, risking up to **35% total harvest loss**. - **Motto:** *One Mission. Build the Impossible.* --- ## 👥 Team: LOGIC LEGION (Contributors & Responsibilities) In compliance with the official **OMNIKON 2026 Hackathon Eligibility & Documentation Guidelines**, the registered team members, their roles, and their key technical ownership areas are documented below: | Contributor | GitHub Username | Role | Key Architectural Ownership | |:---|:---|:---|:---| | **Sahil** | `@Sahil-web01` | **MERN Full-Stack Architect** | **Level 2 Lead & Kisan Ecosystem Architect:** Offline-first Kisan identity & PIN authentication, Zero-bloat personal scan diary, certified digital Kisan Prescription with 1-click WhatsApp dispatch, bilingual Hindi/English localization engine, PWA Service Worker precaching (31 assets, 11 MB), and NGO deferred sync pipeline. | | **Lakshay Gauniyal** | `@Lakshay030105` | **AI & Deep Learning Lead** | **Level 1 Lead:** MobileNetV2 fine-tuning on PlantVillage dataset (38 categories), browser-optimized Float32 TF.js model export (8.922 MiB bundle), 38/38 verified Python-to-TF.js parity validation, secondary Computer Vision edge segmentation filter (OpenCV.js / Canvas) for precise leaf surface area severity grading, and Grad-CAM class activation mapping. | --- ## 🚀 Round 2 Elite Feature …