🏗️ Blueprint Computer Vision Pipeline

This project implements a complete computer vision pipeline tailored for blueprint documents, construction site photos, and technical diagrams. It processes images, runs object detection models (such as YOLOv8), and exposes results through a RESTful API using FastAPI.

---

✅ Features

• 📷 Image preprocessing: resizing, grayscale conversion, histogram equalization
• 🤖 Object detection with modern models (YOLOv8 by Ultralytics)
• 🧪 Postprocessing: extract class labels, bounding boxes, and confidence scores in structured JSON
• 🔌 REST API to send images and receive detection results
• 🧱 Modular design, easy to extend with segmentation, OCR, or PDF parsing

---

📁 Project Structure

blueprint-cv-pipeline/

api/ # FastAPI app for REST API -> main.py

src/ # Computer vision pipeline logic -> preprocess.py # Load and preprocess input images -> inference.py # Run object detection (YOLOv8) -> postprocess.py # Parse and format detection outputs -> pipeline.py # Command-line runner (CLI pipeline)

data/ # Input images (temporary) outputs/ # Annotated results and JSON outputs requirements.txt # Python dependencies README.md # This file Dockerfile # Optional: containerization support