CAD drawing retrieval method and system based on double-branch deep learning
A technology of deep learning and drawing, applied in the field of communication, can solve problems such as lack of recognition efficiency and accuracy
Active Publication Date: 2020-05-08
WUHAN UNIV
View PDF6 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] Most of the existing retrieval and recognition technologies rely on the geometric and topological features of CAD drawings, ignoring the important impact of local salient features on the retrieval and recognition of drawings. In addition, the commonly used image retrieval technology based on deep learning usually takes the entire CAD drawing image as input, directly Obtaining visual features is lacking in recognition efficiency and accuracy
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0036] In order to better understand the technical solution of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.
[0037] see figure 2 , a kind of CAD drawing retrieval method based on deep learning provided by the embodiment of the present invention, comprises the following steps:
[0038] 1) CAD drawing annotation. During the specific implementation, the existing CAD drawing library can be used, or the CAD drawing sample data can be collected in advance, and each drawing data can be marked in advance. Drawing CADrawing i The annotation format is defined as (img i ,Class i ,ID i ), where img i Indicates the serial number of the drawing, the img of each drawing i Unique and unique; ID i Indicates the identity of the entity corresponding to the drawing, an ID i Can correspond to multiple img i ;Class i Indicates the category to which the entity belongs (such as interio...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a CAD (Computer-Aided Design) drawing retrieval method based on double-branch deep learning, which comprises CAD drawing labeling, double-branch deep convolutional network construction, acquisition of a shared layer of shallow depth features of a sample, an inter-class feature extraction branch and an intra-class feature branch; performing loss function setting, ternary testgroup construction and training sample construction, specifically, selecting ternary legend groups with different compositions from a CAD drawing library in a random extraction mode to serve as training samples, and it is guaranteed that at least two legends in each legend group belong to the same object or category; distributing an anchor legend, a positive legend and a negative legend to each group of legends; spatial mapping: training the double-branch deep convolutional network according to the training sample, and mapping a CAD drawing from an original image space to a feature expressionvector space; and for any to-be-retrieved CAD drawing, spatial mapping is realized by utilizing the trained double-branch deep convolutional network, and a matching result is obtained.
Description
technical field [0001] The present invention relates to the field of communication technology, in particular to a CAD drawing retrieval method and system based on dual-branch deep learning. Background technique [0002] Relying on powerful graphics processing capabilities and simple practicality, computer-aided design (CAD) technology has been widely used in various fields such as construction engineering and chemical design. As the core component of CAD technology, CAD drawings are increasingly becoming an important part of engineering development and product design. Faced with a wide variety of CAD drawing data with huge demand, how to quickly and accurately find the design drawings that meet the specific needs of users has gradually become a major problem and challenge for R&D and innovation in various fields. [0003] Most of the existing retrieval and recognition technologies rely on the geometric and topological features of CAD drawings, ignoring the important impact ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/53G06F16/583G06K9/46G06K9/62
CPCG06F16/583G06F16/53G06V10/462G06V10/757G06F18/214
Inventor 何政叶刚王中元傅佑铭
Owner WUHAN UNIV