Semi-automatic extraction of characters from an image into a font for the generation of images and old documents.

This application can be tested online, or installed locally. To install it two solutions are provided : manually or from a docker image.

// Opencv installation
#Ubuntu Linux distribution
$ sudo apt-get install libopencv-dev
    
#MAC distribution (with Homebrew)
$ brew install opencv

// Libnavajo installation
$ git clone https://github.com/titi38/libnavajo.git
$ cd libnavajo
$ cmake .
$ make
$ sudo make install
    
// Font-extractor installation
$ git clone https://github.com/rtisne/Pdp.git
$ cd Pdp/server
$ make

First install docker https://docs.docker.com/engine/installation/

$ docker run -dti -p  80:8080 rtisne/font-extractor

This may take some times. After the installation just go to your docker ip address

To lauch the server, go into the server folder and make:

$ ./main

The server will be run by default on port 8080. You can change the port in the file libnavajo/src/WebServer.cc of libnavajo

#define DEFAULT_HTTP_PORT 8080

If you are in local, you can access to the server throught: http://localhost:8080

The application is host online : https://intranet.iut.u-bordeaux.fr/navajo/

Tests are available to test the algorithm part that retrieves components in the image. To launch test go to the test folder and build:

$ cmake.
$ make
$ ./Unit_test

Developed by:

• Tristan Autret
• Anthony Boschet
• Thomas Labrousse
• Romain Tisné

Report link