DigiSketch: Automating the Conversion of Hand-Drawn Diagrams to Digital

Hand-drawn diagrams play a crucial role in various fields, from education to engineering, yet their integration into the digital realm remains a time consuming and error prone process. This research paper introduces an innovative approach to automatically digitize hand-drawn diagrams, leveraging advancements in computer vision, machine learning, and image processing techniques. Our approach begins with acquiring high-resolution scans or images of hand-drawn diagrams, performing a series of preprocessing steps, enhancing image quality, and correcting distortions to ensure the most accurate representation using annotation tools to annotate diagrams. Further, this work employed TrOCR for text recognition, ensuring precise transcription of handwritten labels. The extracted components are vectorized into scalable and editable digital formats. This approach not only saves time but also preserves the nuances of hand-drawn diagrams, making it understandable to educators and engineers. The automatic digitization of hand-drawn diagrams presented in this paper bridges the gap between analog and digital format, simplifying the transition and enhancing the usability of these valuable visual representations in modern age.