Abstract:As digitization in engineering progressed, circuit diagrams (also referred to as schematics) are typically developed and maintained in computer-aided engineering (CAE) systems, thus allowing for automated verification, simulation and further processing in downstream engineering steps. However, apart from printed legacy schematics, hand-drawn circuit diagrams are still used today in the educational domain, where they serve as an easily accessible mean for trainees and students to learn drawing this type of diagrams. Furthermore, hand-drawn schematics are typically used in examinations due to legal constraints. In order to harness the capabilities of digital circuit representations, automated means for extracting the electrical graph from raster graphics are required.

What problem does this paper attempt to address?

The paper aims to address the problem of automated extraction of hand-drawn circuit diagram images. Specifically, with the advancement of digitization in the engineering field, circuit diagrams (also known as schematics) are typically developed and maintained in computer-aided engineering (CAE) systems for automated verification, simulation, and other downstream engineering processes. However, in the educational field, hand-drawn circuit diagrams are still widely used, especially in teaching and examinations, because they are easy to obtain and convenient for students to learn how to draw such diagrams. Additionally, due to legal restrictions, hand-drawn circuit diagrams are more common in exams. To leverage the advantages of digital circuit representations, automated means are needed to extract electrical diagrams from bitmap graphics. Although some methods have been proposed in the literature, they are usually based on small or non-public datasets. Therefore, this paper describes a modular end-to-end solution and evaluates the methods for various subtasks on a larger, publicly available dataset to form a new benchmark. These subtasks include: 1. **Object Detection**: Used to identify electrical symbols and text. 2. **Binary Segmentation**: Distinguishes strokes from the background. 3. **Handwritten Character Recognition**. 4. **Orientation Regression of Electrical Symbols and Text**. 5. **Computer Vision Graphical Assembly and Correction Algorithms**. All methods are integrated into a publicly available prototype. Through this approach, hand-drawn circuit diagrams can be converted into computer-readable forms, thereby supporting teachers and students in using existing automated verification and simulation tools, improving the speed and objectivity of feedback. Additionally, the method assumes that the input images are snapshots taken under real-world conditions rather than professional scans, making the processing conditions more challenging.

Modular Graph Extraction for Handwritten Circuit Diagram Images

Instance Segmentation Based Graph Extraction for Handwritten Circuit Diagram Images

A Public Ground-Truth Dataset for Handwritten Circuit Diagram Images

Recognition of Hand-Drawn Designs of Electronic Analog Circuits

Automated Recognition of Electrical Circuit Drawings

Graph-Based Locality-Sensitive Circuit Sketch Recognizer

Parsing Netlists of Integrated Circuits from Images via Graph Attention Network

Hand-Drawn Electrical Circuit Recognition Using Object Detection and Node Recognition

Offline hand-drawn circuit component recognition using texture and shape-based features

Hand-drawn Electric Circuit Diagram Understanding Using 2D Dynamic programming

Functional Component Descriptions for Electrical Circuits based on Semantic Technology Reasoning

Component segmentation of engineering drawings using Graph Convolutional Networks

Recognition of symbols and topology in the image of a DC circuit diagram based on contours and skeletons

An Overview of Hand-Drawn Diagram Recognition Methods and Applications

DocEDA: Automated Extraction and Design of Analog Circuits from Documents with Large Language Model

Generation of Netlist from a Hand drawn Circuit through Image Processing and Machine Learning

A two-stage CNN-based hand-drawn electrical and electronic circuit component recognition system

Transforming Engineering Diagrams: A Novel Approach for P&ID Digitization using Transformers

On-line hand-drawn electric circuit diagram recognition using 2D dynamic programming

VectorGraphNET: Graph Attention Networks for Accurate Segmentation of Complex Technical Drawings

Practical single-line diagram recognition based on digital image processing and deep vision models