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

A fault detection method for the round pin of the lower rod of railway freight cars based on deep learning

A railway freight car and deep learning technology, which is applied in image analysis, image enhancement, instruments, etc., can solve the problems of detection accuracy and low detection efficiency, and achieve improved fault identification effect, accurate component segmentation, and rapid convergence Effect

Active Publication Date: 2022-02-01
HARBIN KEJIA GENERAL MECHANICAL & ELECTRICAL CO LTD
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention aims to solve the problems of low detection efficiency and detection accuracy that need to be improved in existing detection and recognition methods

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
  • A fault detection method for the round pin of the lower rod of railway freight cars based on deep learning
  • A fault detection method for the round pin of the lower rod of railway freight cars based on deep learning
  • A fault detection method for the round pin of the lower rod of railway freight cars based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0053] Specific implementation mode one: combine figure 1 To describe this embodiment in detail,

[0054] This embodiment is a deep learning-based method for detecting the failure of a round pin of a lower rod of a railway freight car, including the following steps:

[0055] Obtain the image of the lower rod to be detected, input the image of the lower rod into the segmentation network of the lower rod round pin to predict the round pin of the lower rod and the area of ​​the round hole where the round pin is not installed, and obtain the predicted multi-valued image, 1 value in the multi-valued image is the round pin area, and 2 is the round hole area where the round pin is not installed; judge whether the round pin is lost or not by the number of the round pin area and the round hole area where the round pin is not installed. If the round pin is lost, determine that the round pin is lost s position;

[0056]As long as the round pin area is located, it is necessary to detect...

specific Embodiment approach 2

[0057] This embodiment is a deep learning-based detection method for the round pin fault of the lower rod of a railway freight car. The process of obtaining the image of the lower rod to be detected includes the following steps:

[0058] Obtain the real passing car image as the original image to be detected; according to the wheelbase information and the prior information of the lower rod components, the lower rod is roughly positioned in the original image, and the lower rod image is obtained.

[0059] Other steps and parameters are the same as those in the first embodiment.

specific Embodiment approach 3

[0060] This embodiment is a deep learning-based fault detection method for the round pin of the lower rod of a railway freight car. Before inputting the image of the lower rod into the segmentation network of the round pin of the lower rod for prediction, it is necessary to perform local histogram equalization and enhancement on the lower rod image.

[0061] 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 deep learning-based fault detection method for round pins of lower rods of railway wagons belongs to the field of image detection technology. The present invention aims to solve the problems of low detection efficiency and detection accuracy to be improved in existing detection and recognition methods. In the present invention, the image of the lower rod is input into the segmentation network of the lower rod round pin to predict the round pin of the lower rod and the area of ​​the round hole where the round pin is not installed, and judge whether the round pin is lost. If the round pin is lost, determine the position where the round pin is lost. ;Locate the round pin area, according to the position of the round pin in the segmentation result, intercept the round pin sub-image in the lower rod image, and input the classification network of the lower rod round pin to classify the round pin sub-image; if the round hole is recognized , then determine the position of the round hole where the round pin is not installed; if it is recognized that the cotter pin is missing, then determine the position of the cotter pin loss; if no fault is detected, then process the next image of the pull-down rod. It is mainly used for the fault detection of the round pin of the lower rod of the railway freight car.

Description

technical field [0001] The invention relates to a fault detection method for a round pin of a pull down rod. It belongs to the technical field of image detection. Background technique [0002] If the cotter pin or round pin of the lower rod of the truck is lost, the lower rod will fall off, which will cause a major accident and cause serious losses to personal property. [0003] The falling of the lower rod is one of the most frequent accidents in the use of railway freight cars. In the fault detection of the lower rod cotter pin, the fault detection is basically carried out by manually checking the image, because the inspectors are prone to fatigue and omissions during the work process. And other situations, resulting in the occurrence of missed detection and wrong detection, affecting driving safety. The use of automatic image recognition can not only save a lot of manpower and material resources, but also effectively reduce the influence of human factors, reduce missed ...

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): G06V20/20G06V10/764G06V10/74G06T7/00
CPCG06T7/0008G06T2207/10004G06T2207/30248G06T2207/20081G06T2207/20084G06V20/20G06F18/22G06F18/24
Inventor 刘丹丹
Owner HARBIN KEJIA GENERAL MECHANICAL & ELECTRICAL CO LTD