Steel billet positioning visual recognition and measurement system

Based on intelligent vision, the steel billet online positioning recognition and measurement system installs an automatic recognition system between the furnace outlet and the rolling mill. The system uses AI recognition technology with the help of high-speed cameras to determine the billet position. Positioning solves the problem of steel bar strength and quality problems caused by the billet staying on the roller with too low a temperature entering the rolling mill.


steel billet


Identification requirements and technical requirements

1) Steel head and tail detection: set a feedback point at the No. 1 rolling mill outlet. After the steelhead reaches this position, the steelhead positioning jumps to the second steelhead, and the positioning of the first steel tail is kept outside the safety threshold.

2) Set the safety threshold window: adjust the threshold with the on-site production. When the steelhead and steel tail trigger the safety threshold, the roller stops to prevent collision with steel.

3) Positioning recognition accuracy: less than or equal to 50mm.

4) Dustproof and high temperature resistant, high reliability.

5) Working environment: -15~60℃.

6) System stability.

System composition

The steel billet positioning and identification system is mainly composed of an industrial camera system, visual server system, transmission system, back-end data processing system, and real-time image recording system.

1. Front-end high-speed industrial camera system

Due to the high temperature and humidity at the metallurgical site and the high metal dust content, in addition to the three-proof requirements required by conventional outdoor equipment. The front-end system must also meet the requirements of high-temperature resistance, waterproofness, and anti-interference.

1) The front-end camera system uses a special high-speed and high-temperature-resistant camera with a resolution of up to 12 million pixels. It has automatic shutter speed adjustment and can quickly identify the moving billet.

2) The camera cover has an anti-interference function to deal with the interference of high-frequency equipment and metal dust, reduce the impact of image recognition caused by weather, and reduce manual maintenance work.

3) Meet the national safety protection standard IP66; global scanning CCD sensor; metal shell cover, anti-electromagnetic interference; industrial standard high reliability; 1Gpds Ethernet transmission.

2. Steel billet positioning, identification and measurement software and algorithm

1) Install industrial cameras based on AI recognition technology on-site, computer image visual analysis technology, separate the background and target in the scene, analyze and track the target appearing in the camera scene, judge the position information of the head and tail of the billet in the specified area, communicate with the factory system, and locate the target in the scene. When the behaviour of the predefined rules appears, the system will automatically output a stop signal according to the preset rules and execute the logic to control the stop of the roller.

2) The system software adopts our intelligent and innovative position recognition algorithm and self-learning software system. The system aims at the complex scenes in the production area, improves the traditional Gaussian mixture model, automatically optimizes the parameters, and adds the deep learning algorithm based on the convolutional neural network to ensure the real-time and accuracy of the video analysis. It can automatically identify the target type and track the moving target appearing in the video area, and can automatically capture, identify, analyze, detect and calibrate the billet in real-time and efficiently.

3) Visual recognition processing system host

The visual recognition processing system host is equipped with a high-performance CPU and GPU, which can efficiently realize the operation of artificial intelligence algorithms, greatly improve the system recognition efficiency, and effectively identify high-speed objects.

4. Control module

Siemens PLC is used as the main controller, PROFINET communication is used internally, and the communication with the unit is carried out through PROFINET, and interlocking control is carried out through hard-wired IO.

5. System-related transformation

The Ethernet interface reserved for factory communication supports the PROFINET protocol for integration in the production workshop. PLC single control can also be achieved.

Functional description

1) AI algorithm can automatically monitor whether the steel billet is delivered in place;

2) Automatic identification of crane occlusion and avoidance algorithm.

3) High production efficiency, uninterrupted work, strong adaptability to the environment, and can work in harsh environments such as dust;

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!