PID2Graph

The PID2Graph dataset is a collection of P&IDs (Piping and Instrumentation Diagrams) paired with their corresponding graph structures. This dataset provides opportunities to explore the relationships between symbols and lines in P&IDs, enabling various applications in fields like engineering and computer vision.

Citation

To cite this dataset, please cite the following paper:

Stürmer, J. M., Graumann, M., & Koch, T. (2024)Transforming Engineering Diagrams: A Novel Approach for P&ID Digitization using TransformersarXiv preprint arXiv:2411.13929, https://arxiv.org/abs/2411.13929

Data Structure

In this dataset, symbols are represented as nodes, each with a bounding box defined by its coordinates (xmin, ymin, xmax, ymax) and a label. Lines are represented as edges, also accompanied by a label. The ground truth information is stored in a .graphml file, which can be easily imported and converted into a network graph using libraries like NetworkX in Python.

Note that the graphs include nodes with a "connector" label, which marks the connection point of symbols and lines as well as "border-nodes" for the patched data, that marks the location of a line "leaving" a patch.

The dataset is organized into two main categories:

Complete Plans: This section contains entire P&ID plans, each with its associated graph structure.

Patched Plans: This section contains a collection of patched P&ID plans with overlaps, where each patched file has its own corresponding graphml file. This allows for the analysis of overlapping regions and the relationships between them.

Further information and analysis of the dataset can be found in the paper mentioned above.

Sources

Dataset PID    Referenced by: Paliwal, S., Jain, A., Sharma, M., & Vig, L. (2021). Digitize-PID: Automatic Digitization of Piping and Instrumentation Diagrams. ArXiv, abs/2109.03794.    License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)

PID2Graph OPEN100    Taken from: https://www.open-100.com/    Created by: Energy Impact Center (https://www.energyimpactcenter.org/)

PID2Graph Synthetic    Symbols: Copyright 2023 Steven Baltakatei Sandoval (https://baltakatei.com/)    License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)