Smart Camera Kvantron Smart

Kvantron Smart Camera was developed as a domestic hardware and software platform for replacing foreign industrial code readers and machine vision cameras, used on mandatory marking lines.

The solution is focused on the «Честного знака» challenges: reading marking codes, controlling Data Matrix code application quality, checking readability, event registration and data transmission to traceability systems.

The camera can be used on packaging, aggregation, rejection, shipment control, acceptance, and production accounting lines.

The smart camera automatically detects the code in the frame, pre-processes the image and identifies the identifier even in difficult production conditions: with high line speed, unstable lighting, vibrations, product position changes or insufficient contrast of the marking.

The system can control:

• code presence on the product or packaging;
• correct positioning of the code;
• readability and contrast;
• damage, blur, incomplete print or code deformation;
• code compliance with production process requirements;
• fact of successful or unsuccessful reading;
• transmitting result to an external system.

The camera works with codes on labels, caps, boxes, vials, bags, parts, product bodies and other industrial goods.

Apart from code recognition, Kvantron smart camera can be used as a universal visual quality control module.

The system analyzes the product image and detects deviations from the standard:

• surface defects;
• spots, contaminants, chips, cracks;
• shape or geometry violations;
• incorrect item position;
• missing elements;
• assembly errors;
• color or shade deviations;
• printing and label defects;
• packaging completeness violations;
• product non-compliance with required specification.

For AI-control tasks, computer vision, machine learning and neural object / defect classification algorithms are used. Brightness transformations, filtration, distortion compensation, binarization, mathematical morphology, image statistics, as well as YOLO and U-Net neural architectures are applied in Kvantron Smart.

The key feature of Kvantron smart camera is PLC-based hardware acceleration for image processing.

PLC allows some operations to run directly inside the device:

• initial image processing;
• noise filtering;
• frame normalization;
• boundary and contour detection;
• region of interest search;
• preparing image for code recognition;
• faster camera response to line events.

This approach reduces processing lag and enables use on high-speed lines, where fast decision making is essential.

The smart camera operates on the unified logic of the Kvantron Smart software platform. The platform is responsible for scenario setup, image processing, event visualization, result storage, and data exchange with external systems.

Kvantron Smart enables:

• creating recognition and control scenarios;
• setting decision rules;
• configuring quality thresholds;
• managing cameras and lighting;
• saving event and defect images;
• generating statistics by shifts and lots;
• transferring data to marking and traceability systems;
• integration with line equipment and actuators.