2025-26

Permanent URI for this collection

http://192.168.75.5:4000/handle/123456789/21836

Browse

Browsing 2025-26 by Author "ARPITA KUMARI 1NH24MC020"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    AN AI-Powered full-Stack Mobile Application for Plant Identification and Health Monitoring
    (NHCE, 2026) ARPITA KUMARI 1NH24MC020
    This report presents the design and development of FloraFlow, an AI-powered, full-stack mobile application engineered to automate the plant care lifecycle and optimize urban smart gardening. Developed during a full-stack developer internship at Emergent Labs, the platform replaces highly inefficient manual cultivation methods and subjective guesswork with a data-driven digital ecosystem. The system architecture utilizes a modern, decoupled framework consisting of a React Native (Expo) cross-platform frontend client and an asynchronous FastAPI (Python) backend. Multi-tenant data privacy and secure user onboarding are established via Google OAuth 2.0, routing data directly to an elastic MongoDB Atlas NoSQL cloud database. The core innovation of the application is its AI Intelligence Module, which integrates Vision-to-Text Large Language Models (LLMs). When a user captures a plant photograph, the frontend converts the compressed asset into a Base64-encoded ASCII string. The backend processes this payload concurrently and proxies it to the AI vision engine, which parses complex biological markers such as leaf venation and necrotic decay to return structured JSON diagnostic summaries, scientific species identification, and target care remedies. To sustain plant vitality, a decentralized Smart Reminder Module leverages the expo-notifications framework to offload background tasks directly to the device's hardware kernel, calculating precise watering windows using chronological epoch arithmetic. Finally, a Garden Analytics Module builds a real-time data aggregation pipeline on the user's profile dashboard, summarizing overall garden health into immediate Healthy, Attention, or Critical status cards. Ultimately, FloraFlow successfully demonstrates how merging advanced computer vision, non-blocking asynchronous architectures, and flexible document-oriented databases can transform a manual, error-prone routine into a structured, automated hobby.