How shubhamsharmayo/image-segmentation works
Project type and purpose
• This is a full-stack AI-powered web application.
• Frontend: React single-page application (SPA) that provides UI for object detection, video processing, webcam streaming, and handwriting recognition.
• Backend: Flask REST API that orchestrates CV/AI pipelines (Mask R-CNN via OpenCV DNN, PyTesseract OCR).
File structure
• Top-level:
• client/ → all React/Tailwind code (components, assets, routing)
• server/ → Flask app (app.py), model files (dnn/, coco.names), (uploads/)
• Patterns: clear separation of concerns (UI vs. API).
• No microservices: a monolithic Flask backend.
• Build & config: client/package.json, vite.config.js, tailwind.config.js, server/requirements.txt.
README for additional insights
• Tech stack table confirms React + Flask + OpenCV + Tesseract + TensorFlow model.
• Lists API endpoints, so you know which UI screens connect to which backend routes.
• Uploads flow: client sends image/video, backend saves to /uploads, processes, returns result.
Final advice
• Be exhaustive: map each React component to its API endpoint.
• Model load sequence (e.g., “load frozen_inference_graph_coco.pb on startup”).
• Call out any configuration files (vite.config.js, tailwind.config.js, eslintrc).