Digitization of Chess Board and Prediction of Next Move

In recent years digital technologies and gadgets have grown so much. Along with other things, the field of Artificial Intelligence has also grown dramatically. Computer vision is one of the most compelling types of AI, which everyone must have experienced in one or the other way like in OCR, Image Recognition, object detection, google lens, etc. One might even come to the assumption that given any image or video to a computer, the computer can understand what’s going on in it and can comment a few things on it, just like WE HUMANS. But that assumption is not yet very true. It may be possible in the future but at the present these need some work to become reality. Leveraging this same motivation, we intend to give some contribution for our dream future to become reality. Recognizing and understanding any arbitrary chess board from the image, thereby digitizing it, is the thing that we will be achieving from the proposed system. Along with all that, giving a comment on the current state of chess board, thereby predicting the next optimal move in the game is another thing which is included in proposed system. Think of a human-like robot trying to play chess. Robots must have something like this in order to play the game. In the proposed research work digitization of Chess Board is done by server application from chess board image acquired by user. Current state of the board is generated and sent back as a response to client application. With the help of GUI, the chessboard is generated from the received Forsyth-Edwards Notation (FEN) on the client side. Depending upon blacks turn or whites turn next optimal move to be played is shown to the user. With our proposed system digital instance is generated with accuracy of over 90%. All blank cells of the board are generated with almost 100% accuracy.