Engineering Drawing Retrieval System Based on Bag of Visual Words Model

Engineering Drawing Retrieval System Based on Bag of Visual Words Model

연구 분야: Analysis

학회: 2022 IEEE 6th Advanced Information Technology, Electronic and Automation Control Conference (IAEAC )

Aiming at the reuse of existing CAD engineering drawings in manufacturing product design, a 2D engineering drawing retrieval method based on bag of visual words model is proposed. The scale invariant feature transformation (SIFT) algorithm is used to extract visual vocabulary vectors from different kinds of 2D engineering drawings, and the K-Means algorithm is used to cluster the extracted SIFT features to form visual vocabulary, construct visual dictionary, construct the bag of visual words model of 2D engineering drawings, and realize the retrieval of engineering drawings. Select 1000 engineering drawings, build bag of visual words model of engineering drawings, and build a 2D engineering drawing retrieval system. The method of retrieving engineering drawings using product point cloud data is proposed. The point cloud data of enterprise products is scanned by 3D scanner to generate multi views of STL model., extract the contour of product view, extract the characteristics of contour view, and retrieve similar 2D engineering drawings, so as to provide help for reverse engineering.