This is an OpenCV-based Python chess bot that automates the playing of chess on a computer. The bot detects the chessboard on the screen using an efficient algorithm and then leverages UCI chess engines such as Stockfish or Leela to make moves. The bot uses PyAutoGUI library to interact with the chessboard and make moves. It captures screenshots of the chessboard using d3dshot library on Windows, and MSS library on Linux. Additionally, there is also a remote version of the bot that uses a cloud server to process moves and then sends them to the client.

The bot requires the following libraries to be installed:

1. OpenCV
2. PyAutoGUI
3. NumPy
4. d3dshot (Windows only)
5. MSS (Linux only)
6. python-chess
7. python-socketio-client
8. You can install these libraries by running the following command:

$ pip install opencv-python pyautogui numpy d3dshot python-chess

1. Make sure the chess engine you want to use (e.g., Stockfish or Leela). It resides in the same folder as bot-offline/online.py does
2. Make sure the chess board is visible on the screen.
3. Start a cmd terminal by it's side and try:

   python chess_bot.py

4. Bot will automatically detect the board, the color you are going to play and will start making the moves
5. You can edit the parameters of the bot i.e. threads, ram, time & depth from default.ini settings file
6. Since the promotion to the pieces differ you can tweak a bit of the code to promote to any piece and give the signal back to work so that it continues playing.

The bot uses OpenCV to detect the chessboard on the screen. It first captures a screenshot of the screen using d3dshot or MSS, depending on the operating system. It then applies a series of image processing techniques to detect the corners of the chessboard.

Once the corners are detected, the bot uses PyAutoGUI to interact with the chessboard and make moves. It first captures a screenshot of the chessboard and then uses computer vision techniques to detect the pieces on the board. It then uses the python-chess library to generate a list of legal moves and selects the best move based on the evaluation of the chess engine.

In the remote version of the bot, the client sends the screenshot of the chessboard to the cloud server, which then processes the image and sends back the best move to the client.

Watch the following video to see the bot in action:

This Python code is designed to run as a server on a cloud computer with an open port. The server takes a chessboard position in Forsyth–Edwards Notation (FEN) as input and returns the best move back to the client using sockets. This feature is useful in offloading the computational load from the client, allowing the bot to run on old devices without any performance issues.

The code uses the socket module to create a socket object and bind it to a specific address and port. It also uses the chess and chess.engine modules to calculate the best move. The requests module is used to fetch data from the cloud API(Endgame Tablebase), and the json module is used to decode the JSON response.

The no_of_pieces variable is used to count the number of pieces on the board. If there are less than 8 pieces on the board, the code fetches the best move from the cloud API. Otherwise, it uses the engine object to calculate the best move based on the given time control or depth.

Finally, the server object listens for incoming client connections and spawns a new thread to handle each connection. The threading module is used to create a new thread for each client connection. The code prints information about each client connection, including the client's IP address and the total number of active connections.

Overall, this code provides a simple and efficient way to run a chess bot on a cloud server and offload the computational load from the client.

This OpenCV-based Python chess bot provides a simple yet effective way to automate the playing of chess on a computer. It is easy to use and supports any UCI chess engine. Additionally, the remote version of the bot allows users to offload the computation to a cloud server, making it possible to run

• Add chessbooks to save time/resources initially.
• Add Server support and processing
• Add piece recognition algorithms(AI/ML) to make playing more robust
• Improve code quality.