http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
객체 위치를 기반으로 한 P&ID 연결 정보 인식 연구
정두영(DooYoung Jeong),박광필(Kwang-Phil Park),임다빈(Dabin Lim),정현(Hyun Chung) (사)한국CDE학회 2022 한국CDE학회 논문집 Vol.27 No.4
In the piping layout design process, it takes a lot of hours to read the piping and instrumentation diagram (P&ID) and find the connection information. P&IDs are often delivered as images or non-digital drawings for various reasons. In this study, to extract connection information automatically from drawing images, we recognize text, symbols, and lines, and organize connection relation based on the objects’ location. Tesseract OCR, an optical character recognition engine, was used for text recognition, and an Object Detection model and OpenCV were used in combination for symbol and line, and those are implemented in Python. Connection between the objects detected was evaluated based on the coordinates of them and some pixel distance considering the drawing resolution. To verify the proposed method, the possibility of extracting drawing information was reviewed by testing 9 P&ID drawings provided by a shipyard. The overall recognition rate was 74.8% for text and more than 97% for symbols. A digital file which includes the connection relation between the pipes and the instruments generated after the recognition stage, and it shows that the method finds the connection between pipe-pipe and pipe-instrument in the accuracy of 76.2% and 89.7% respectively.