AI solutions for the steel industry

Application of machine vision in strip steel surface defect detection

Strip steel is the main product of metallurgical enterprises and the main raw material for automobile, national defence, chemical industry and light industry manufacturing. The global energy shortage and the rising price of iron ore have put forward higher requirements for the yield, qualification rate and high-quality rate of strip steel. Surface quality is the main quality indicator of strip steel. Due to factors such as raw materials, rolling process, system control, etc., defects such as cracks, scars, roll marks, holes, skin delamination, stains, pitting, etc. may appear on the surface of strip steel. It’s inevitable. These defects have varying degrees of impact on the fatigue resistance, corrosion resistance, wear resistance, and electromagnetic properties of the strip. Therefore, during the rolling process, these defects should be discovered in time, the control parameters should be adjusted, and strips of different grades should be classified and marked. , has far-reaching significance for improving the quality of strip steel products, reducing production energy consumption, and maximizing sales profits.

At present, in strip production, manual inspection methods are basically used to check defects on the strip surface. This method has many shortcomings: it is easily affected by the subjective factors of the inspector, and the inspection results lack accuracy, reliability and completeness; the inspection of the lower surface must turn over the plate, thus reducing productivity; the labour intensity of workers is high; when the temperature of the steel plate When it is high, it is difficult for quality inspectors to stand very close to the steel plate, thus affecting the detection effect.

 

Cold rolled strip steel surface inspection system

 

The system mainly consists of four parts, including a detection box, on-site control box, data processing cabinet and operation console.

(1) Detection box:

The detection box is equipped with two high-speed linear array CCD cameras, and high-intensity parallel light sources, and is equipped with installation and adjustment mechanisms, protection systems, etc. There are wheels under the detection box, and a guide rail positioning system is installed on the foundation. The detection box can be pulled out from the production line. , easy for inspection and maintenance. The function of the detection box is to obtain image data of the strip surface.

(2) On-site control box:

The on-site control box is installed near the detection box to complete the electrical control in the detection box, collect image data, and transmit the preprocessed data and equipment status information to the data processing cabinet. The strip speed pulse signal provided on-site enters the on-site control box and cooperates with the camera acquisition and control.

(3) Data processing cabinet:

The data processing cabinet is installed in the operating room and consists of multiple data processors and servers. The data processor processes the image data of the steel plate in parallel, transmits the processing results to the server for comprehensive analysis and statistics, and saves the data.

(4) Operation console:

The operation console is installed in the operation room and connected to the server to perform various operations.

 

Detection principle

 

The cold rolling online surface defect detection system uses the principle of optical imaging detection to realize strip surface defect detection by using high-speed linear array CCD cameras, linear parallel light sources, high-performance data processing systems and the latest machine vision image processing technology.

Cam1 and Cam2 are high-speed linear array CCD cameras, and L is a linear parallel light source. The light source emits high-intensity and stable parallel light, which is projected onto the steel plate under test at a certain angle, forming a narrow light band in the width direction of the steel plate.

The Cam1 camera (called a brightfield camera) is installed on the reflection channel of the light source and receives the light reflected from the steel plate. If there is a defect on the steel plate and the reflective properties of the defective part change, the camera will capture this change; the brightfield image has Characteristics of light background and dark defects.

The Cam2 camera (called a dark field camera) is installed between the light source and the bright field camera and receives the light diffusely reflected by the steel plate. If there are no defects on the steel plate, the light obtained by the dark field camera is very weak. When there are defects on the surface of the steel plate, because defects increase the diffuse reflection of light, the image captured by the dark field camera will also change. The dark field image has the characteristics of a dark background and bright defects. In this way, the frame-by-frame images output by the camera at a certain shooting speed can be spliced into a complete strip surface image. These image data are sent to the image processing computer to identify defects on the strip surface.

Using two cameras, brightfield and darkfield, most defects can be detected by complementing each other, such as epidermal delamination, scales, rust spots, adhesion, pits, etc. with brightfield defect characteristics, and cracks, etc. with darkfield defect characteristics. Creases, holes, pinholes, etc.

Cold-rolled strip surface quality monitoring is developed based on the rapid development of machine vision and computers. It makes full use of the most advanced image acquisition equipment and image processing algorithms. The key technologies include image acquisition and image data processing.

(1) Image acquisition

The acquisition of images is related to the light source and the installation position of the camera. The light source uses a high-brightness LED red light source, and the light emitted is parallel light. The camera uses a high-speed, high-sensitivity camera, and the installation position is at a certain angle with the light emitted by the light source. When there are no defects on the surface of the steel plate, the light hitting the steel plate is reflected out, and the image obtained by the camera is darker; when there are defects on the surface of the steel plate because the uneven surface of the defect will produce diffuse reflection, at this time the camera will get a brighter image at the corresponding sensor position. In this way, the images frame by frame are combined (coordinated with the strip speed) to obtain equidistant surface image data of the entire strip.

(2) Data processing

The image recognition algorithm is complex and the amount of data processing is large, so multi-processor parallel processing technology is used. After each processor is processed, it is sent to the server, which is organized and saved by the server. The location and type of the defect are recorded, and the location and type of the defect are recorded. Partial images are saved to disk, and various defects are analyzed and compared through the expert system and the defect library to obtain statistical data, which can be used to guide the optimization process. It can identify common roller marks, scratches, rust marks, feather marks, bonding, folding marks, etc. The system can gradually improve the defect recognition rate through continuous learning.

 

steel coil, strip steel, defect detection

The cold-rolled strip surface quality monitoring system is a non-contact testing equipment developed based on the principle of machine vision. It has the characteristics of easy installation, low maintenance, and high-cost performance. After installing the system, it can provide objective product quality analysis reports and quickly respond to production. problems, thereby guiding the optimization of production processes, improving production efficiency, reducing scrap rates, and reducing quality costs.

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!