CLens: An AI-Based Framework for Converting Handwritten Chess Scoresheets into PGN

Abstract

This research paper introduces CLens, an AI-powered system developed to convert handwritten chess scoresheets into a structured digital format. In many offline chess tournaments and practice sessions, players record moves manually on paper, which makes later review and storage difficult. The proposed system aims to simplify this process by automatically transforming handwritten move records into a machine-readable format that can be used for digital analysis. The system combines image processing, deep learning–based handwriting recognition, and chess move validation to achieve accurate results. First, the image of the scoresheet is processed using computer vision techniques to identify and extract the handwritten move entries. These extracted moves are then passed through a trained deep learning model that interprets the handwriting and converts it into standard chess notation. To improve correctness, the generated moves are checked against legal chess rules so that invalid or impossible moves can be detected and corrected. Finally, the validated move sequence is converted into Portable Game Notation (PGN) format, allowing users to replay the complete match and perform detailed game analysis on digital chess platforms. By integrating artificial intelligence with rule-based validation, CLens provides an efficient and practical solution for bridging traditional handwritten chess recording with modern digital analysis tools.