Circuit Vision: Fast and Efficient Digitization of Complex Hand-Drawn Circuit using Yolov8

In this research, we propose an innovative method for automating the digitization of hand-drawn circuit diagrams by utilizing the YOLOv8 object detection framework. Hand-drawn circuits are commonly utilized in educational contexts and prototyping scenarios; however, converting them into digital formats for analysis and simulation is often laborious and prone to errors. Our approach aims to address this challenge by leveraging advanced deep-learning techniques. We employ the YOLOv8 model to detect and precisely locate circuit components such as resistors, capacitors, transistors, integrated circuits, and wires within hand-drawn circuit diagrams. The framework was thoroughly trained and refined using various hand-drawn circuit datasets. We achieve robust performance in accurately identifying and delineating individual circuit elements, even amidst variations in drawing styles and quality. After component detection, we implement a series of post-processing procedures to reconstruct the circuit topology and extract pertinent connectivity information. This includes methodologies such as line tracing for wire direction identification, component classification, and spatial relationship analysis for component localization. The resultant digitized circuit representation faithfully preserves the original layout while offering a structured and machine-readable format suitable for further analysis and simulation.