ya-ota/TrainingAssistant

A tool for creating Open-CV training samples

Set up

1. Clone to your computer:

    % git clone git@github.com:shkh/TrainingAssistant.git
   

2. Add Jcrop:

    % cd TrainingAssistant
    % git submodule init
    % git submodule update
    % cd static/Jcrop
    % git checkout master
   

3. Install python modules

    % pip install -r freezed.txt
   

4. Add images into static/img

Run server

% python views.py

This command starts the Flask server on port 5000, visit http://localhost:5000.

After all images will be processed, you will get info.dat and bg.txt; Respectively the list of positive samples and the list of negative samples.