Method for efficiently identifying component primitives based on eigenvalue analysis

The invention relates to the technical field of component primitive recognition in DWG drawings, in particular to an efficient component primitive recognition method based on eigenvalue analysis. A feature value analysis-based component primitive efficient identification method comprises the following stages: 1, construction of a charting rule reading original data of a DWG drawing, including the type, coordinate, attribute and other information of a primitive; 2, constructing a component object recognition rule base; 3, DWG drawing processing; 4, component feature value extraction; and 5, machine learning training and identification. Through the steps of machine learning training and recognition, automatic and intelligent recognition and classification of the components in the drawing can be achieved, errors and low efficiency of manual recognition are avoided, and the recognition accuracy and efficiency are improved. 本发明涉及DWG图纸中构件图元识别技术领域，尤其涉及一种基于特征值分析的高效识别构件图元方法。基于特征值分析的高效识别构件图元方法，包括以下各阶段：一、制图规则库构建：读取DWG图纸