Deep learning defect recognition method based on multi-feature fusion
A multi-feature fusion and deep learning technology, applied in the field of deep learning defect recognition based on multi-feature fusion, can solve the problem of high detection accuracy, achieve high recognition efficiency, realize automatic and intelligent recognition, and good accuracy.
Active Publication Date: 2021-05-18
易思维(杭州)科技股份有限公司
View PDF12 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] In order to solve the above technical problems, the present invention provides a deep learning defect identification method based on multi-feature fusion. This method is based on deep learning, without manual intervention in the whole process, and can intelligently classify and identify defects, with high detection 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
[0064] The technical solution of the present invention will be described in detail below in conjunction with specific embodiments.
[0065] A deep learning defect recognition method based on multi-feature fusion, used to detect surface defects of mirror / mirror-like objects, using the display screen to project multiple sinusoidal stripes sequentially on the surface of mirror / mirror-like objects; the camera collects and projects on the mirror / mirror-like objects multiple sinusoidal fringes on the surface, and record the collected multiple sinusoidal fringe images as a sinusoidal fringe atlas (such as figure 2 As shown, they are the atlases of sinusoidal fringes collected corresponding to the gray-black defect, wear mark defect, and natural-color scratch defect (when using one frequency and four phases));
[0066] Use the sinusoidal fringe atlas to identify surface defects on specular / mirror-like objects, such as figure 1 shown, including the following steps:
[0067] 1) using...
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 discloses a deep learning defect recognition method based on multi-feature fusion, which collects a sinusoidal fringe atlas, respectively obtains at least two kinds of distribution diagrams in a phase distribution diagram, a curvature distribution diagram and a gray scale distribution diagram, and records them It is a feature atlas; the feature atlas is input into a pre-trained deep learning model, the probability of each defect type corresponding to the feature atlas is obtained through the processing of the deep learning model, and the mirror surface / class is identified according to the size of the probability The type of mirror object surface; this method obtains multiple types of feature distribution maps for different types of defect images at the same time. Compared with traditional methods, it has the characteristics of high recognition efficiency and good accuracy, and can realize the automation and intelligence of mirror defect categories identify.
Description
technical field [0001] The invention relates to the field of defect detection, in particular to a deep learning defect recognition method based on multi-feature fusion. Background technique [0002] Mirror / mirror-like objects widely exist in modern manufacturing industries, such as automobile painted bodies, aircraft painted bodies, electronic display panels, optical mirrors, polishing molds, etc.; the surface quality of mirrored objects is affected by processing technology, coating quality, manufacturing Influenced by factors such as the environment, some defects will inevitably occur, such as pits, bulges, scratches, dirt, etc. The existence of defects will not only affect the reflection, transmission, anti-corrosion and other functions of mirror objects, but also reduce their surface aesthetics. Therefore, it is necessary to detect the surface defects of mirror / mirror-like objects in time, so as to ensure that the mirror objects can meet the high standard requirements of...
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 Patents(China)
IPC IPC(8): G06T7/00G06K9/62G06K9/46
CPCG06T7/0004G06T2207/20081G06T2207/20084G06T2207/20076G06V10/454G06F18/24G06F18/253G06F18/214
Inventor 尹仕斌郭寅孙博赵进
Owner 易思维(杭州)科技股份有限公司