阅读说明：本技术 一种基于机器视觉的钢坯宽度在线检测方法及系统 (Machine vision-based billet width online detection method and system ) 是由 冯小雷 崔忠信 田绍鹏 田兴 刘海峰 于 2021-08-26 设计创作，主要内容包括：本发明涉及一种基于机器视觉的钢坯宽度在线检测方法及系统,系统包括依次连接的激光器、控制器、计算机,激光器垂直安装于轧钢辊道上方,发射出条状蓝色激光束,激光束在被测钢坯上发生散射,反射光被激光器的光电接收装置接收并成像,将图像转变为电信号；利用计算机上的视觉图像处理软件,将获得的图像转化为256灰度级的浓淡图像,在钢坯轮廓图样上取采样点,采样点a-(1)b-(1),a-(2)b-(2),a-(3)b-(3)……a-(n)b-(n)；利用视觉图像处理软件测量各个采样点的值；选取各个采样点测量值中的最大值作为钢坯宽度测量值。本发明自动化程度高,智能高效,能够快速、准确的检测轧机辊道上钢坯的宽度值,为生产提供指导。(The invention relates to a machine vision-based billet width online detection method and a machine vision-based billet width online detection system, wherein the system comprises a laser, a controller and a computer which are connected in sequence, the laser is vertically arranged above a steel rolling roller way and emits strip-shaped blue laser beams, the laser beams are scattered on a billet to be detected, reflected light is received and imaged by a photoelectric receiving device of the laser, and the image is converted into an electric signal; converting the obtained image into a 256-gray-level gray-scale image by using visual image processing software on a computer, and taking a sampling point a on the outline pattern of the billet 1 b 1 ，a 2 b 2 ，a 3 b 3 ……a n b n (ii) a By using visionMeasuring the value of each sampling point by the image processing software; and selecting the maximum value in the measured values of the sampling points as the measured value of the width of the steel billet. The method has the advantages of high automation degree, intelligence and high efficiency, and can quickly and accurately detect the width value of the steel billet on the roller way of the rolling mill, thereby providing guidance for production.)

1. A billet width online detection method based on machine vision is characterized by comprising the following steps:

s1, acquiring billet images by using a billet image acquisition device arranged above the roller table;

s2, transmitting the acquired billet image to a computer through a controller;

s3, extracting a plurality of sampling points longitudinally in the billet image by using processing software on a computer;

s4, obtaining the width value of each sampling point;

s5, analyzing and processing the obtained numerical values of the sampling points, and removing numerical values with obvious differences;

and S6, obtaining the width measurement value of the steel billet for production control.

2. The on-line detection method for the width of the steel billet based on the machine vision as claimed in claim 1, characterized in that: in S1, the billet image acquisition device is a two-dimensional laser, and the acquisition mode is continuous acquisition.

3. The on-line detection method for the width of the steel billet based on the machine vision as claimed in claim 2, characterized in that: in S3, 3-8 pairs of sampling points are extracted randomly along the length direction of the billet.

4. The on-line detection method for the width of the steel billet based on the machine vision as claimed in claim 3, characterized in that: in S4, a sub-pixel edge location algorithm is used to calculate the width value of each sampling point.

5. The on-line detection method for the width of the steel billet based on the machine vision as claimed in claim 4, characterized in that: in S5, the removal of the significant difference value refers to the removal of the head and tail of the billet, and the significant excess value due to the error of the sampling point.

6. The on-line detection method for the width of the steel billet based on the machine vision as claimed in claim 5, characterized in that: in S6, the maximum value of each sampling point is selected as the final width measurement value.

7. The utility model provides a steel billet width on-line measuring system based on machine vision which characterized in that: the detection method comprises a billet image acquisition device, a controller electrically connected with the billet image acquisition device and a computer in communication connection with the controller, wherein the billet image acquisition device is arranged above a roller table and is vertical to the running direction of a billet.

8. The on-line billet width detection system based on machine vision as claimed in claim 7, wherein: the billet image acquisition device is a laser; the controller is connected with the laser through a shielding signal line, and the computer is connected with the controller through the Ethernet.

9. The on-line billet width detection system based on machine vision as claimed in claim 8, wherein: the laser is a two-dimensional laser, and the two-dimensional laser adopts a continuous acquisition mode.

Technical Field

The application belongs to the technical field of automatic detection, and particularly relates to a machine vision-based billet width online detection method and system, which can realize online measurement of billet width.

Background

In the steel rolling production process, the guide plates on two sides of the rolling mill are required to be used for centering the steel billet, so that the steel billet is prevented from deviating, the width of the steel billet needs to be measured, and the guide plates are accurately controlled. In the hot rolling production process, the temperature of the steel billet is about 1100 ℃, the width of the steel billet can be measured only by adopting a non-contact measurement method, and in addition, the width of the steel billet needs to be measured on line in real time due to the broadening of the steel billet in the rolling process.

Disclosure of Invention

The invention provides a machine vision-based billet width online detection method and system for the precise control of a guide plate of a rolling mill according to the actual operation condition of a billet in the steel rolling process, and belongs to the field of online non-contact measurement.

In order to solve the problems, the technical scheme adopted by the invention is as follows:

a billet width online detection method based on machine vision comprises the following steps:

s1, acquiring billet images by using a billet image acquisition device arranged above the roller table;

s2, transmitting the acquired billet image to a computer through a controller;

s3, extracting a plurality of sampling points longitudinally in the billet image by using processing software on a computer;

s4, obtaining the width value of each sampling point;

s5, analyzing and processing the obtained numerical values of the sampling points, and removing numerical values with obvious differences;

and S6, obtaining the width measurement value of the steel billet for production control.

The technical scheme of the invention is further improved as follows: in S1, the billet image acquisition device is a two-dimensional laser, and the acquisition mode is continuous acquisition.

The technical scheme of the invention is further improved as follows: in S3, 3-8 pairs of sampling points are extracted randomly along the length direction of the billet.

The technical scheme of the invention is further improved as follows: in S4, a sub-pixel edge location algorithm is used to calculate the width value of each sampling point.

The technical scheme of the invention is further improved as follows: in S5, removing the significantly different values means removing too small values and significantly too large values due to errors in sampling points.

The technical scheme of the invention is further improved as follows: and S6, selecting the maximum value of the sampling points meeting the requirement as the final width measurement value.

A machine vision-based billet width online detection system is used for realizing the detection method and comprises a billet image acquisition device, a controller electrically connected with the billet image acquisition device and a computer in communication connection with the controller, wherein the billet image acquisition device is arranged above a roller table and is vertical to the running direction of a billet.

The technical scheme of the invention is further improved as follows: the billet image acquisition device is a laser; the controller is connected with the laser through a shielding signal line, and the computer is connected with the controller through the Ethernet.

The technical scheme of the invention is further improved as follows: the laser is a two-dimensional laser, and the two-dimensional laser adopts a continuous acquisition mode.

Due to the adoption of the technical scheme, the invention has the beneficial effects that:

the method and the system have high automation degree, belong to the category of on-line non-contact measurement, are simple to build and assemble, are intelligent and efficient, can quickly and accurately detect the width value of the steel billet on the roller way of the rolling mill, provide a theoretical basis for the width control of the guide plate, are used for production, and have higher application value.

Drawings

FIG. 1 is a schematic illustration of the structural installation of the system of the present invention;

FIG. 2 is a schematic representation of a billet image sampling point of the present invention;

wherein: 1. the device comprises a billet image acquisition device, 2, a controller, 3, a computer, 4, a roller table, 5 and a billet.

Detailed Description

The present invention will be described in further detail with reference to examples.

The invention discloses a machine vision-based billet width online detection system, which comprises a billet image acquisition device 1, a controller 2 electrically connected with the billet image acquisition device 1, and a computer 3 in communication connection with the controller 2, namely the billet image acquisition device 1, the controller 2 and the computer 3 are sequentially connected with one another, referring to fig. 1.

As can be seen from fig. 1, the billet image capturing device 1 is firmly installed above the roll table 4 and perpendicular to the running direction of the billet 5. In the specific application connection, the billet image acquisition device 1 is a laser, so that the controller 2 is connected with the laser 1 through a shielding signal line, and the computer 3 is connected with the controller 2 through the Ethernet. The laser acquisition mode is continuous acquisition, so that the accuracy of acquiring a sample can be ensured.

Preferably, the laser is a two-dimensional laser, the two-dimensional laser adopts a continuous acquisition mode, and the test precision is higher.

The invention also discloses a billet width online detection method by utilizing the detection system, which comprises the following steps:

s1, billet image acquisition is carried out by using the billet image acquisition device 1 arranged above the roller table 4;

s2, transmitting the acquired billet image to the computer 3 through the controller 2;

s3, extracting a plurality of sampling points longitudinally on the billet image by using processing software on the computer 3;

s4, obtaining the width value of each sampling point;

s5, analyzing and processing the obtained values of the sampling points, and removing obviously different values, such as too small values;

and S6, obtaining the width measurement value of the steel billet for production control.

In step S1, the billet image acquisition device 1 is a laser, the laser emits a blue stripe laser beam with a wavelength of 405nm, the laser beam is scattered on the billet 5 to be measured, the reflected light is received and imaged by the photoelectric receiving device of the laser, and the image is converted into an electrical signal by using a CMOS device.

In step S2, the controller 2 processes, filters and amplifies the electrical signal, and transmits the signal to the computer 3 via the ethernet. The controller 2 can be a conventional chip, such as an STM microprocessor, a microchip chip, etc.

In step S3, the billet contour pattern obtained by the visual image processing software on the computer is a plan view of the billet 5 as shown in fig. 2. Sampling points are taken from the billet image, and the sampling point a is shown in figure 2₁b₁，a₂b₂，a₃b₃……a_nb_nAnd comprises a plurality of pairs of sampling points in the longitudinal direction. The image processing software can be HALCON, MVIMS, MVICS and the like, and does not have excessive requirements.

Step S4, converting the image into a 256-gray-scale image through the visual image processing software, and generating an image most suitable for measurement, which can improve the stability of measurement. And calculating the numerical value of each sampling point by adopting a sub-pixel edge positioning algorithm. The sub-pixel edge location algorithm is prior art and will not be described herein.

Step S5, as can be seen from fig. 2, n pairs of sampling points are selected and respectively located at the head, the tail and the billet body of the billet 5, and there are multiple pairs of sampling points on the billet body, where the sampling point a₁b₁And a_nb_nThe width values of the head and the tail of the billet 5 are obviously lower and obviously different, the billet is removed, and sampling points on the billet body are reserved. Meanwhile, obviously overlarge numerical values obtained by sampling point acquisition errors are removed.

For other reserved sampling points meeting the requirement, the maximum value is selected as the final width measurement value because the final width value is used for controlling the width of the guide plate. Or taking the median or average of the values of the reserved sampling points, multiplying by a certain proportionality coefficient k, such as k = 1.1-1.6, and finally multiplying by the final product to obtain a final width calculation value, and hooking the final width calculation value with the width of the guide plate, thereby also achieving the purpose of the invention.

And step S6, selecting the maximum value of each sampling point meeting the requirement as the final width measurement value. After the width measured value of the steel billet is obtained, the computer 3 transmits the value to the industrial field PLC through the Ethernet, and the industrial field PLC adjusts the width of the guide plate according to the width value of the steel billet.

The width measuring device is convenient to use, high in automation degree, intelligent and efficient, and capable of rapidly and accurately detecting the width value of the steel billet on the roller way of the rolling mill, and the online real-time measuring function of the width of the steel billet is greatly improved.