README.md

Signature Detection App

A web application for signature detection.

It's based on the signature detection tool I created.

Features

This app has 3 features.

• detect the signature in an image

• correct the error in the detection result

• add a signature to the result

In the end, user can download the result as a json file.

Detect the signature

Correct the error

Add a signature

Install

This application is shipped with Docker.

After downloading the codes, you can run it in 2 modes.

dev mode:

cd signature_detection_app
docker-compose up

production mode:

cd signature_detection_app
docker-compose -f docker-compose.yml up

That's all! The application is run at localhost.

Structure

The application is composed by 3 parts.

• frontend

  It is created by CRA (create-reat-app), written by Typescript.

• backend

  It uses Fastapi, a Python web framework.

• web server

  It is Nginx. It proxys the requests.

For now, no database is connected. It's very easy to extend with docker-compose.

Testing

• Frontend:

  CRA includes the testing-library.

  To test the file you create or modify, run:

  cd signature_detection_app/frontend
  npm test -- --coverage

  To test all the files, run:

  cd signature_detection_app/frontend
  npm test -- --watchAll=false --coverage

• backend:

  I uses pytest and coverage for backend testing.

  To see the test coverage, run:

  cd signature_detection_app
  source backend-coverage.sh 

Future reading

• Innovation project: from Lab to Production

  It's an article I wrote about how this project is created.

• signature_detection Github repo

  It's a Github repo I created that explains the signature detection algorithm. It also includes the Jupyter notebooks to play with.

• signature-detect package

  It's a PyPI package of signature detection. It can detect the signature from both images and PDF files.