Research on the Analysis and Check of Electrical Secondary PDF Drawings Based on Deep Learning

摘要

Aiming at the problems such as the large amount of manual checking, the influence of the quality of checking on the technical level of the personnel, the quantity of the project and the working attitude, and the influence of the quality of the drawings on the progress and quality of the engineering construction, a method for analyzing and checking PDF drawings based on deep learning was proposed. Using artificial intelligence and image processing technology to analyze and identify drawings, and designing automatic checking of drawings, the quality and efficiency of drawing checking were improved. Among them, the analysis method of drawings mainly includes the conversion of PDF files into picture files, the identification of components in the drawings, the identification and screening of lines, the segmentation of long lines according to the location of components, the segmentation of line according to the relationship between lines and the output of wires and components. It is of great significance to quickly analyze and study the PDF drawings. To some extent, it can avoid the non-standard situation after drawing, improve the quality of engineering design, and provide a strong guarantee for the follow-up checking work.