<p>Flow charts are often used in the business world for managing, documenting, designing, and analyzing transactions and applications in various fields. These schemes contribute to the regulation of processes in enterprises and increase productivity. Although these diagrams can be drawn with the help of applications that facilitate drawing in electronic environments, they can also be drawn manually on surfaces such as papers, both in meetings and during drafting. The problem of digitizing arises when these drawings are intended to be developed or stored. In this study, a dataset was created by making drawings on papers and taking photographs of these drawing. The flow diagram drawn by hand on paper is transferred to a digital platform that allows changes to be made on it. During this transformation, the You Only Look Once (YOLO) deep learning model that can be used in the field of object classification and detection was used. The digitized flowchart is also represented in a form of Extensible Markup Language (XML), so that it can be used by any other application and on any other platforms. For the test purposes, the Drawio platform was used.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Digitization of hand-drawn flow charts with deep learning

  • Ahmet Sayar,
  • Ahmet Fatih Mustacoglu

摘要

Flow charts are often used in the business world for managing, documenting, designing, and analyzing transactions and applications in various fields. These schemes contribute to the regulation of processes in enterprises and increase productivity. Although these diagrams can be drawn with the help of applications that facilitate drawing in electronic environments, they can also be drawn manually on surfaces such as papers, both in meetings and during drafting. The problem of digitizing arises when these drawings are intended to be developed or stored. In this study, a dataset was created by making drawings on papers and taking photographs of these drawing. The flow diagram drawn by hand on paper is transferred to a digital platform that allows changes to be made on it. During this transformation, the You Only Look Once (YOLO) deep learning model that can be used in the field of object classification and detection was used. The digitized flowchart is also represented in a form of Extensible Markup Language (XML), so that it can be used by any other application and on any other platforms. For the test purposes, the Drawio platform was used.