Defect detection of “rail surface” based on machine vision

With its incomparable advantages in transportation speed, transportation capacity and transportation cost, railway transportation occupies a pivotal position in the transportation industry.

As railway transportation is moving towards the goal of fast transportation speed and heavy cargo load, in order to ensure the absolute safety of passenger travel, the quality requirements for railway transportation components are also getting higher and higher, and the surface quality of rails is raised. More stringent requirements, in which surface defects are an important technical index to measure the surface quality of rails.


1. Difficulties in the detection of rail surface defects


At present, the problems existing in the detection of rail surface defects are as follows:

(1) Low degree of intelligence

Through on-the-spot investigation, it was found that the existing large-scale steel rail and beam factories such as Panzhihua Iron and Steel Co., Ltd. and Baotou Steel still use manual visual inspection to monitor the surface quality of the rails. The production efficiency is low, and there is little reference value for subsequent process improvement.

(2) Less research on rail defect detection

At present, the relatively mature defect detection technology of steel products mainly focuses on steel plates, and there is little research on cold rails.

(3) The surface material of the rail is special and difficult to handle

Rail is a highly reflective material, and its surface grey level does not change much. Therefore, rail defect detection has higher requirements for imaging quality and defect segmentation algorithms.


2. Analysis of typical defects on the rail surface


By analyzing the actual situation in the production process of an iron and steel enterprise and the formation mechanism of rail defects, the surface defects of cold rails can be divided into three typical defects: surface inclusions, rolling scars, and rolling marks. In addition, there is other kinds of defects that are similar to defects but not Defects pseudo-defects.
The appearance of these types of defects is as follows:


Steel trails surface defect


(1) Surface inclusions

Cause of formation: Non-metallic inclusions and slag inclusions on the surface of the rail are broken during cold rolling
Features: Irregular in size and shape, sheet-like folds, complex texture, off-white or brown-red, etc.

(2) rolling scar

Causes: Most of them are caused by severe wear and tear during the rail production process or metal foreign objects left by incomplete descaling

Features: Irregular outline, mostly “concave” shape, distributed on the side of the rail

(3) Roll marks

Causes: The roll is severely worn or the iron sheet is stuck on the roll, etc.
Features: Often straight or arc-shaped grooves, varying in depth, relatively slender compared to other defects

(4) False defects

Causes of formation: Mostly caused by the peeling off of the iron oxide scale during the washing process
Features: large area, large length, uniform texture inside the defect, often brownish red


3. System hardware structure design


Light source selection


Steel rails have the following properties: strong reflective properties, uniform material, and frequent changes in the light at the production site. Therefore, it is very important to choose a light source with adjustable illuminance and uniform illumination for rail imaging.

In the bright field lighting mode, most of the reflected light on the surface of the rail enters the camera, while the diffuse reflection of the defect makes the light not return in the same way, so it is relatively dark relative to the background, and the reflectivity of the material on the rail surface makes the resulting image easy to blur. exposure;

In the dark field lighting mode, part of the light enters the camera due to diffuse reflection at the defect position, so it is brighter than the background, so as to achieve the purpose of highlighting the defect.

Considering that the common defects on the rail surface are mostly concave shapes such as cracks, rolling scars, and inclusions, dark-field lighting is used.

In the lighting scheme adopted by the system, the line-scan camera is placed vertically above the rail, and the light source obliquely illuminates the surface of the rail.


Camera selection


In the actual production line, the measured rail moves at a speed of 1.5m/s, and the running speed is relatively fast. The frame rate of the area array camera cannot meet the requirements, and redundant imaging is easy. In addition, the image acquisition speed of the CMOS camera is fast, and it is more suitable for shooting moving objects, so it is more suitable for this system.


4. Rail surface defect detection system


The purpose of the detection system is to detect and identify rail surface defects, so the design of the detection system should meet the following requirements:


(1) Ensure the complete imaging of the rail surface


Due to the side-to-side fluctuation of the rail during actual operation, in order to ensure the complete imaging of the rail surface, we must reserve enough imaging width to meet the demand.


(2) Ensure that rail defects can be displayed normally


In the actual production line, the running speed of the cold rail is 1.5m/s, so the imaging speed of the built imaging system should be consistent with the running speed of the rail.


(3) Defect detection category: can automatically detect the surface of the rail


Realize the identification and classification of rolling scars, rolling marks, inclusions and pseudo defects.


(4) Defect detection rate and recognition rate: higher than 90%


The defect detection rate refers to the percentage of the number of detected defects to the total number of defects, and the recognition accuracy refers to the percentage of the number of correctly identified defects to the total number of defects

The designed detection system mainly consists of three modules: image acquisition, image processing and motion control.
Among them, the image acquisition module and the motion control module belong to the hardware system, and the image processing module belongs to the software system.

In this detection system, the image acquisition module is composed of a camera, a lens and a light source, and is placed above the rail to be tested, with adjustable installation height and angle; the motion control module is used to transport the rail and control the sampling frequency of the camera; the image processing module is mainly used to The camera collects rail surface images for processing and realizes the detection and identification of rail surface defects.

Leave a Reply

Your email address will not be published. Required fields are marked *

With machine vision as the core, DBM can provide intelligent equipment R&D, manufacturing, and transformation services such as unmanned factories in the metallurgical industry and other industries, machine vision, and robot application design.

Product category

Customer testimonials

Tell us your needs and a technician will contact you immediately.

Friendship Link

We are committed to becoming a “leader in visual applications”, advocating the core values of “integrity, cooperation, innovation, expertise, and excellence”, taking the road of “technology shaping the brand, innovation leading the industry”, and attaching importance to the investment in scientific research and technology. The research and development of new products and the construction of the team make unremitting efforts for the goal of scientific and technological development.

We have been supplied to world’s top steel manufacturer Arcelormittal, TATA Steel, EZZ steel etc. We do OEM for Concast and Danieli for a long time

Metallurgical Intelligence
Change the way of manufacturing

Save labor, improve working environment, improve efficiency, reduce error rate and save cost!

Металлургический интеллект
Изменить способ производства

Сэкономьте труд, улучшите рабочую среду, повысьте эффективность, уменьшите количество ошибок и снизьте затраты!