Products
Vehicle Make & Model Recognition (MMR)
Recognition of 16 vehicle classes & 17 road user categories
Our MMR recognizes 16 classes of vehicles & 17 road user categories. Recognition of vehicles covers 336 makes, 2,722 models, and 7,122 generations—including variations and colors. Recognition of road users covers vehicles, bicycles, pedestrians, motorcycles, scooters, e-scooters, kickbikes, kick scooters, tricycles, quadbikes, trailers, trolleys, prams, luggage, animals, rickshaws, and pedestrians pushing trolleys.
Try it for free at https://trafficeye.ai, or contact filip.kachnic@eyedea.ai
Number Plate Reading (ANPR)
License plate recognition
Detects all plates in images or video, including special characters, states, single/multi-line formats, ADR codes, vehicle sets, and wheel counts. Designed for accuracy-critical mobility sectors, with MMR offering market-leading vehicle recognition. Trusted by camera makers, e-tolling providers, integrators, and agencies like INTERPOL and Europol. Accuracy exceeds 99% even in rain, fog, snow, or tailgating scenarios.
Test for free at https://trafficeye.ai/ or contact filip.kachnic@eyedea.ai
Traffic Analyzer
AI-Powered Traffic Analysis from Images, Video Footage, and Live Streams
A desktop application for traffic data collection that integrates our MMR and ANPR systems. The application reads license plates and independently recognizes other vehicle attributes such as make, model, generation, vehicle color, & direction of view. Field-proven solution grounded in real-world traffic conditions.
The console version is available as a Docker image and can be executed in parallel as many times as needed. The output data can be directed to an FTP server.
Forensic ANPR
Reading unreadbale license plates & vehicle classification
A desktop application for offline analysis of unreadable license plates in degraded video or photo evidence. It uses a proprietary deep neural network trained specifically for recovering license plate content from degraded visual input, such as motion blur, low resolution, poor lighting, or oblique angles. Equipped with our MMR, it enables recognition over 15,000 make-model-generation-variant combinations. It currently supports 36 countries.
For trial contact filip.kachnic@eyedea.ai
Vehicle Cabin Visual Inspection
Image and Video AI Traffic Surveillance
With years of experience in AI recognition for traffic, we are also able to deliver solutions such as Vehicle Cabin Visual Inspection. We can process the low-resolution vehicle cabin images in low light conditions and limited view through the windshield. This allows us to detect seat occupancy, seat belts fastened and driver distraction.
For trial contact filip.kachnic@eyedea.ai
TrafficEye Cloud
World-leading Number Plate Recognition (ANPR) fused with most detailed Make & Model Recognition (MMR)on the market.
Anonymization
Anonymizing human faces, vehicle license plates and vehicles.
Eyedea’s Anonymizer is a high-speed, high-precision AI-powered tool for anonymizing human faces and vehicle license plates in images and videos. It ensures privacy by automatically detecting and blurring sensitive elements, even in low-resolution or spherical street-view inputs. Built on the WaldBoost algorithm and trained with massive datasets, it’s offered as an SDK or GUI for Windows and Linux, supporting both CPU and GPU processing.
UIC wagon number reading
UIC wagon number recognition software system
Eyedea’s OCR system reads complete UIC wagon numbers from side-view images or video, even on moving wagons. It works independently of wagon type or construction, with >99% accuracy across Europe. Custom hardware integration, high adaptability, and confidence scoring per character make it ideal for railway operators, infrastructure firms, and authorities. Software can be prepared for other sensor types. Windows 64-bit. Perpetual license + 12-month support.