Unlock instant, AI-driven research and patent intelligence for your innovation.

Defect sample synthesis method, defect inspection network training method, computer readable medium, computing system, and image inspection device

A synthesis method and defect inspection technology, applied in computing, image analysis, image data processing, etc., can solve problems such as lack of simulation ability, inability to retain the original background of the image, and inability to simulate interaction

Pending Publication Date: 2022-07-01
DELTA ELECTRONICS INTL SINGAPORE
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, neither of the above two methods of data augmentation can preserve the original background of the image
Therefore, although the existing solutions can generate defects and their backgrounds at the same time, they cannot simulate the interaction between defects and normal backgrounds due to the lack of simulation capabilities, that is, they cannot simulate defects in different backgrounds.
The existing method of synthesizing defect samples cannot provide unknown information for the defect inspection model, so it is not an effective method to use the existing method of synthesizing defect samples to amplify data to improve the defect inspection model based on convolutional neural network

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Defect sample synthesis method, defect inspection network training method, computer readable medium, computing system, and image inspection device
  • Defect sample synthesis method, defect inspection network training method, computer readable medium, computing system, and image inspection device
  • Defect sample synthesis method, defect inspection network training method, computer readable medium, computing system, and image inspection device

Examples

Experimental program
Comparison scheme
Effect test

preparation example Construction

[0046] figure 1 is a flow chart of a method for synthesizing defective samples according to an embodiment of the present case, figure 2 is suitable for figure 1 System schematic diagram of the defect sample synthesis method. like figure 1 and figure 2 As shown, the method for synthesizing a defective sample in this application includes the following steps. In step S100, a normal sample 1 and a synthesizer 2 are provided. Next, in step S200 , the normal sample 1 is input to the synthesizer 2 . Then, in step S300, the synthesizer 2 is used to generate the synthesized defect sample 3. By inputting the normal samples 1 to the synthesizer 2 and causing the synthesizer 2 to generate synthetic defect samples 3, real and diverse synthetic defect samples 3 can be generated for training an accurate and widely applicable defect inspection network. In step S400 , the synthesized defect sample 3 is input to the synthesizer 2 . In step S500 , the synthesizer 2 is used to generate t...

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

No PUM Login to View More

Abstract

The invention relates to a defect sample synthesis method, a defect inspection network training method, a computer readable medium, a computing system and an image inspection device. The defective sample synthesis method includes the steps of providing a normal sample and a synthesizer, inputting the normal sample to the synthesizer, generating a synthesized defective sample using the synthesizer, inputting the synthesized defective sample to the synthesizer, and generating a repaired normal sample using the synthesizer.

Description

technical field [0001] This case is about a method for synthesizing defect samples, especially a method for synthesizing defect samples, a method for training a defect inspection network, a computer-readable medium, a computing system, and an image inspection device. Background technique [0002] Deep Convolutional Neural Networks (DCNN) have a wide range of applications in the field of computer vision, such as image recognition and object detection, and the model based on the deep convolutional neural network DCNN has better performance in the above applications. However, the training of the above types of models requires a large amount of labeled data, and since defect samples are very rare and difficult to collect, the application of deep learning-based models in industrial automatic defect inspection is limited, whereas normal samples are usually more abundant and easier to collect. [0003] When defective samples are scarce, data augmentation is a widely used technique...

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
no application Login to View More
IPC IPC(8): G06T7/00G06N3/04G06N3/08
CPCG06T7/0002G06N3/08G06T2207/20081G06T2207/20084G06N3/045
Inventor 张功杰崔凯文吕士健洪慈憶
Owner DELTA ELECTRONICS INTL SINGAPORE