Automatically resolve RidderMaster based on TensorFlow & OpenCV

基于 TensorFlowOpenCV 实现的全自动化解御迷士小马谜题

GitHub GitHub last commit GitHub Repo stars



How to use

Deploy the server yourself

  1. Clone the repo

    git clone
  2. Ensure your local environment install pip3:

    sudo python3
  3. Install dependencies by pip3

    pip3 install -r requirements.txt
  4. Try to run python3 to see whether the dependency is successfully installed (optional)

    WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/tensorflow/python/compat/ disable_resource_variables (from tensorflow.
    python.ops.variable_scope) is deprecated and will be removed in a future version.
    Instructions for updating:
    non-resource variables are not supported in the long term
    WARNING:tensorflow:From load (from tensorflow.python.saved_model.loader_impl) is deprecated and will be removed in a future version.
    Instructions for updating:
    This function will only be available through the v1 compatibility library as tf.compat.v1.saved_model.loader.load or tf.compat.v1.saved_model.load. Th
    ere will be a new function for importing SavedModels in Tensorflow 2.0.

    The correct output should be like this, pay attention to the character A at the end, this is the answer to test.jpg.

On Windows Server, you also need to install this dependency. Visual C++ Redistributable

  1. Run api server by uvicorn or gunicorn If you are personal use, then please run directly.


    If you have a lot of needs, you can through gunicorn.

    gunicorn -c api:app

    The default port is 80.

    Port can be modified in line 53 of or line 13 of #6

  2. Add your server address to the connet list of user.js, and modify the API_SERVER below.

    // @connect
    // ==/UserScript==
    const API_SERVER = '';

    Change to your address, note that there is no / character at the end.

  3. Open our test site and observe if your script is running normally.

Use the server we provide

We cannot guarantee the network quality and stable operation of the server, and it is very likely that the answer will not be returned within the specified time.

Install user.js directly and you can use it. Use to see if it works.