Windshield impact deformation fault identification and detection method, storage medium and equipment

A technology for fault identification and detection methods, applied in neural learning methods, character and pattern recognition, instruments, etc., can solve problems such as the detection accuracy rate needs to be improved, the detection efficiency is low, and the detection accuracy rate is low, so as to improve the stability and reliability. Accuracy, good detection effect, the effect of shortening the time

Active Publication Date: 2022-02-01
HARBIN KEJIA GENERAL MECHANICAL & ELECTRICAL CO LTD
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the problems of low detection accuracy and low detection efficiency in the existing manual image inspection method, and the problem that the detection accuracy of the existing nerve retention over-detection method needs to be improved

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
  • Windshield impact deformation fault identification and detection method, storage medium and equipment
  • Windshield impact deformation fault identification and detection method, storage medium and equipment
  • Windshield impact deformation fault identification and detection method, storage medium and equipment

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0035] This embodiment is a windshield impact deformation fault identification and detection method, including the following steps:

[0036] Obtain an image containing the windshield to be detected, and input the image into the windshield impact deformation fault recognition network for fault detection;

[0037] The windshield strike deformation fault recognition network is a combined network of the improved Trident network and the residual network. The combination of the improved Trident network and the residual network is not directly spliced ​​with the Trident network, but changes part of the structure of the Trident network. , in the structure of windshield impact deformation fault recognition network:

[0038] The three branches of the combined network have the same structure, and for each of the three branches, the output of the last level of the residual network of the branch a l+1′ , first passed to the global average pooling layer, then passed through the first fully...

specific Embodiment approach 2

[0041] This embodiment is a windshield impact deformation fault identification and detection method. In this embodiment, the soft threshold processing is to delete the features whose absolute value is smaller than the threshold, and shrink the features whose absolute value is greater than the threshold toward zero:

[0042]

[0043] Among them, x represents the input amount, corresponding to the output a of the last level of residual network l+1′ ; τ represents the threshold, which corresponds to the threshold obtained after processing the Sigmoid activation function; y represents the result of the soft threshold processing.

[0044] Other steps and parameters are the same as those in the first embodiment.

specific Embodiment approach 3

[0045] This embodiment is a method for identifying and detecting windshield impact deformation faults. In this embodiment, the process of obtaining an image containing a windshield to be detected includes the following steps:

[0046] Obtain the train image, and intercept the image containing the windshield to be detected from the train image.

[0047]Other steps and parameters are the same as those in Embodiment 1 or 2.

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

A windshield impact deformation fault identification and detection method, storage medium and equipment belong to the technical field of image detection. In order to solve the problems of low detection accuracy and low detection efficiency in the existing manual image inspection method. The present invention utilizes the image input windshield hitting deformation fault recognition network to detect the fault of the image to be detected; the windshield hitting deformation fault recognition network adopts the combined network of Trident and residual network, and in each branch of the combined network, the last level of residual network the output of a l+1 , passed to the GAP layer and the FC layer, and then passed to the ReLU activation function, and then passed through a FC layer, passed to the activation function to obtain the threshold; the threshold and the output of the last level of residual network are soft-thresholded; at the same time, the activation The output result of the function is branched, and it is normalized through an FC layer and a Softmax activation layer, and then combined with a l+1 Multiply to get the branch detection result; add the detection results of the three branches to get the final output. Fault identification for windshield strike deformation.

Description

technical field [0001] The invention belongs to the technical field of image detection, and in particular relates to a fault detection method for windshield impact deformation. Background technique [0002] The windshield is a component installed between two adjacent carriages at the end of the train vehicle, used to connect the carriages and form a passage for passengers to enter and exit. Once the windshield is damaged and deformed by foreign objects, its internal stress will change and its structural stability will decrease, which may endanger personal safety and cause heavy losses. In order to ensure the smooth and safe operation of the train, it is necessary to identify and detect the windshield. Once any damage or deformation is found, it needs to be dealt with immediately. [0003] At present, manual inspection of images is generally used to inspect windshields for faults, but inspectors are prone to fatigue and omissions during work, resulting in missed inspections ...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06V10/764G06V10/774G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/214G06F18/241
Inventor 王璐
Owner HARBIN KEJIA GENERAL MECHANICAL & ELECTRICAL CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap