Tunnelwatch is an Automatic Incident Detection (AID) software which detection models are based on the latest Deep Learning and Computer Vision technologies. By filtering the major part of the false alarms formerly registered by the traditional system, it widely improves the accuracy of real-time reported information for tunnel operators.
Based on a leading-edge expertise in Deep Learning, Tunnelwatch ensure a very with high detection rate while filtering false alarms
Thanks to an automatic lane segmentation AI model, camera masks are automatically generated, with the right area classification associated
Compatible with most cameras, Tunnelwatch has no specific equipment constraints. This scheme facilitates system maintainability while enabling limitless evolution
Our Tunnelwatch algorithms have been already trained on millions of videos coming from different tunnels. This robust knowledge base makes it easy to adapt to every context
Automatic lane segmentation, on the set-up phase as well as over time, that ensures a mask configuration adaptive maintenance
AI can recognize object shapes, correlated situations and incident patterns, by taking into account the overall context parameters such as space, time and simultaneity
Thanks to the tracking technology, each vehicle is uniquely identified from one camera to another, enabling a double verification of alarms and display prioritization
Modern and easy-to-use interface at the hand of tunnel supervisors and integrators, to boost system qualification and supervision efficiency
Thanks to a leading-edge expertise in the latest Deep Learning technologies, our detection algorithms have been trained on millions of images to be able to recognize specific shapes and situation patterns. We used AI to strengthen, clarify and synthetize incident detection thanks to an important analytical process: tracking, behavioral analytics, object detection, etc.
We also has developed an AI video-based solution that can detect Transport of Dangerous Goods licence plates, from a normal camera only. Based on a powerful image processing algorithm, this cutting-edge technology is able to reliably and accurately track the number of Dangerous Goods trucks circulating on the road network. This innovative plug-and-play solution enables a more effective monitoring of these specific vehicles that require special attention from the operators.