Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Modular Graph Extraction for Handwritten Circuit Diagram Images

Johannes Bayer, Leo van Waveren, Andreas Dengel

TL;DR

This work tackles the extraction of electrical graphs from hand-drawn schematics captured in real-world images, addressing the lack of public benchmarks for handwritten diagrams. It proposes a modular end-to-end pipeline comprising object detection, binary segmentation, orientation regression, text recognition, edge extraction, and post-processing, all validated on the public CGHD dataset. Key contributions include establishing a reproducible baseline on a large public dataset, detailing dataset annotations (bounding boxes, orientation, text, segmentation, polygons), and delivering a publicly available prototype for reconstructing circuit graphs from raster images. The results demonstrate the viability of modular graph extraction for educational and engineering workflows, with public data/code enabling further research and practical deployment in CAE pipelines and tutoring systems.

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. While respective approaches have been proposed in literature, they are typically conducted on small or non-disclosed datasets. This paper describes a modular end-to-end solution on a larger, public dataset, in which approaches for the individual sub-tasks are evaluated to form a new baseline. These sub-tasks include object detection (for electrical symbols and texts), binary segmentation (drafter's stroke vs. background), handwritten character recognition and orientation regression for electrical symbols and texts. Furthermore, computer-vision graph assembly and rectification algorithms are presented. All methods are integrated in a publicly available prototype.

Modular Graph Extraction for Handwritten Circuit Diagram Images

TL;DR

This work tackles the extraction of electrical graphs from hand-drawn schematics captured in real-world images, addressing the lack of public benchmarks for handwritten diagrams. It proposes a modular end-to-end pipeline comprising object detection, binary segmentation, orientation regression, text recognition, edge extraction, and post-processing, all validated on the public CGHD dataset. Key contributions include establishing a reproducible baseline on a large public dataset, detailing dataset annotations (bounding boxes, orientation, text, segmentation, polygons), and delivering a publicly available prototype for reconstructing circuit graphs from raster images. The results demonstrate the viability of modular graph extraction for educational and engineering workflows, with public data/code enabling further research and practical deployment in CAE pipelines and tutoring systems.

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. While respective approaches have been proposed in literature, they are typically conducted on small or non-disclosed datasets. This paper describes a modular end-to-end solution on a larger, public dataset, in which approaches for the individual sub-tasks are evaluated to form a new baseline. These sub-tasks include object detection (for electrical symbols and texts), binary segmentation (drafter's stroke vs. background), handwritten character recognition and orientation regression for electrical symbols and texts. Furthermore, computer-vision graph assembly and rectification algorithms are presented. All methods are integrated in a publicly available prototype.
Paper Structure (25 sections, 9 figures, 1 table)

This paper contains 25 sections, 9 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. The CGHD Dataset
  4. Images
  5. Bounding Box Annotations
  6. Symbol and Text Orientation
  7. Text Strings
  8. Segmentation Maps
  9. Polygon Annotation
  10. User Interface
  11. Methodology
  12. Object detection
  13. Binary segmentation
  14. Orientation Regression
  15. Text Recognition
  16. ...and 10 more sections

Figures (9)

  • Figure 1: A semantic segmentation map can be synthesized from individual parts of the CGHD dataset.
  • Figure 2: Circuitgraph User Interface.
  • Figure 3: Selected Symbol Templates used as Reference for Orientation Prediction.
  • Figure 4: Learning Curve for Object Detection
  • Figure 5: Learning Curves for Orientation Prediction
  • ...and 4 more figures