Steel rail defect detection method and system based on deep residual shrinkage network
A defect detection and residual technology, applied in neural learning methods, biological neural network models, image data processing, etc., can solve the problems of lagging detection technology, low damage detection rate, high false alarm rate, and achieve high classification accuracy. High rate and robustness, high damage detection rate, and the effect of reducing labor costs
Pending Publication Date: 2022-01-04
SHENZHEN BEYEBE NETWORK TECH CO LTD
View PDF0 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0002] On the basis that the hardware of large rail flaw detection vehicles has gradually matured, the rail damage defect detection technology based on deep learning technology is relatively lagging behind, resulting in low damage detection rate, high false alarm rate, high manual intervention rate, and high labor intensity. It is easy to miss related problems such as rail damage and defects
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
[0039] The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.
[0040] Such as figure 1 As shown, the present invention uses the damage feature extraction and classification processing of the rail damage image data, and finally obtains the damage defect list. The present invention includes the following steps:
[0041] S1: Perform data preprocessing based on the damage image data of existing rails, and clean the data;
[0042] S2: Perform feature extraction on the cleaned damage image data to obtain a corresponding damage feature vector;
[0043] S3: Train and learn the features of the damaged image to obtain the optimal classification model;
[0044] S4: Obtain the rail damage image data, and obtain a list of damage defects after classification by the optimal classification model.
[0045] In step S1 of this example, the preprocessing methods of the rail damage defect image data are: image sharpening, ...
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 steel rail defect detection method and system based on a deep residual shrinkage network, and belongs to the technical field of steel rail defect detection. The detection method comprises the following steps of S1, performing data preprocessing based on flaw image data of the existing steel rail, and cleaning the data; S2, performing feature extraction on the cleaned flaw image data to obtain feature vectors of corresponding flaws; S3, performing training learning on the features of the flaw image to obtain an optimal classification model; and S4, acquiring steel rail damage image data, and classifying the steel rail damage image data through the optimal classification model to obtain a damage defect list. The invention has the beneficial effects that manual intervention is not needed, the labor cost is reduced, and the injury detection rate is high.
Description
technical field [0001] The invention relates to a rail detection method, in particular to a rail defect detection method and system based on a deep residual shrinkage network. Background technique [0002] On the basis that the hardware of large rail flaw detection vehicles has gradually matured, the rail damage defect detection technology based on deep learning technology is relatively lagging behind, resulting in low damage detection rate, high false alarm rate, high manual intervention rate, and high labor intensity. It is easy to miss related problems such as rail damage and defects. Contents of the invention [0003] In order to solve the problems existing in the detection of rail damage defects in the prior art, the present invention provides a rail defect detection method and system based on a deep residual shrinkage network. The rail quality evaluation is carried out through deep learning technology, which can detect more accurately and efficiently Rail damage def...
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 IPC IPC(8): G06T7/00G06N3/04G06N3/08G06K9/62
CPCG06T7/0004G06N3/08G06T2207/20081G06T2207/20084G06T2207/30136G06N3/048G06N3/045G06F18/24G06F18/214
Inventor 黄友文杜卫红谢立欧陈炯魏伟航
Owner SHENZHEN BEYEBE NETWORK TECH CO LTD