Skip to content

Computer vision app to import drawings and images as architecture models

Notifications You must be signed in to change notification settings

lishuaiphd/computer-vision-model-import

Repository files navigation

Overview

This project implements a web app where users upload a photo/scan of a hand-drawn software/system/process diagram. The app extracts shapes, arrows, and text, builds a structured graph, and exports a JSON file representing the graphical elements.

The pipeline uses robust classical Computer Vision (OpenCV) plus OCR (Tesseract). An optional neural segmentation model scaffold is included for improved results (PyTorch).

Setup

  1. Install Tesseract.
  • Linux: sudo apt install tesseract-ocr
  • MacOS: brew install tesseract
  • Windows: install from the internet and add to PATH.
  1. Install all required packages.

pip install -r requirements.txt

Usage

  • Linux/MacOS: bash run.sh
  • Windows: run_windows.bat

Open http://127.0.0.1:5000

About

Computer vision app to import drawings and images as architecture models

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •