Video control system for monitoring presence of personnel and left unattended objects in the area of inspection table at a metallurgical enterprise

In March 2024 Mallenom Systems has completed the implementation of the EYECONT video analytics system at one of the major domestic metallurgical production plants.

The primary objective of the system implementation was to create safe working environment for the personnel by creating timely notifications in cases of industrial health and safety violations and preventing industrial accidents.

System functionality:

• Detection of personnel in the high-risk areas of the inspection surface when performing surface quality control of the rolled metal sheets
• Counting the number of people in the area of the inspection surface for checking compliance with safety regulation requirements
• Control of the left objects (e.g., work tools, etc.) in the area of metal sheets inspection
• Logging of occurred safety violations
• Notification of the responsible bodies in a case of safety violations

System operation principle:

As a part of completed project, the video analytics system has been implemented as a supplementary tool to the main safety and security system at the production site.
The video control system recognises cases when people and left foreign objects present in the area of an inspection surface which is intended for visual quality control of the top and bottom surfaces of the rolled metal sheets. The entry of personnel in the inspection area is only permitted when all of the moving components of the equipment at the inspection site are stationary.
When system detects person during inspection area’ machinery operation, and also in cases when foreign objects have been left by personnel, it generates a signal which is sent to the safety and security system of the enterprise. Also, the delivered system sends notification about detected safety violation as an email to a responsible body.
Furthermore, the video analytics system has the capability for connecting to external loud speakers and light notification system that helps to inform about presence of people during equipment operation.

Project features:

• 4 IP cameras are utilised – with two of each being located at the opposite sides of the inspection area. This ensures complete coverage of the inspection site helping to mitigate any blind spots.
• Additional service was developed that helps to aggregate information from several cameras and display precise number of people in the inspection area helping to deal with data input duplicates.
• Due to substantial inspection area size and relatively small left objects in a camera frame in comparison, innovative neural networks algorithms have been implemented and their adaptation to the unique implementation environment and optical layout has been performed. As a result, the system is capable to recognise objects with minimal resolution of 60x65 pixels.
• The system prevents false identification of inspection area’s working machinery and its components as dangerous events which helps to prevent false system activations.
• EYECONT video analytics system has been integrated with enterprise safety and security system via TCP protocol.
• The system reliably operates in challenging environment of the metallurgical enterprise which include extreme temperatures at both ends of a spectrum, fluctuations in humidity, presence of strong vibrations and high concentration of dust in the air.

For additional information related to the system implementation, please feel free to contact:

Andrey Kalinyak
International Development Manager
Email: sales@mallenom.ru