Video surveillance acts as the workhorse of traffic sensing in the traffic management process. However, in low-light conditions such as foggy and rainy weather, video surveillance will be difficult to function around the clock because of “unclear” and “unclear”. By detecting the reflection frequency of the target, the radar can not only measure the distance, but also measure the speed and detect the real-time position of the target, effectively making up for the lack of video surveillance and allowing the perception system to continue to work.
In addition, the radar detection range is wider than that of video surveillance, which can better play the role of early warning.
The integration of video surveillance and radar realizes the complementary advantages of these two technical solutions to the greatest extent, and opens up a new idea for upgrading roadside perception facilities.
The all-in-one machine of Uniview Technology’s LeTV traffic event detection under Qianfang Technology integrates 80GHz millimeter-wave radar, high-definition AI camera unit, and LED warm light. It is deeply integrated in multiple dimensions such as hardware structure, data acquisition, and data processing to achieve all-weather , over-the-horizon, high-precision perception, widely used in digital roads, smart intersections, holographic tunnels, blind spot warnings, speeding reminders, vehicle-road coordination and other scenarios.
In terms of target recognition, it supports capturing motor vehicles, non-motor vehicles, pedestrians and other targets, and supports recognition of blue, yellow, black, white, green, yellow-green two-color, gradient green and other license plates. In terms of flow detection, it supports queuing length, traffic flow, average vehicle speed, average vehicle length, headway, lane occupancy rate and real-time channel vehicle statistics, and can classify large, medium and small vehicles.
In terms of event detection, it supports event detection functions such as illegal lane changes, abnormal parking, congestion detection, retrograde detection, low-speed alarm, over-speed alarm, queue overflow, etc.
① Wide-area multi-target detection and precise perception
The device can track 256 targets, support 20-250 meters long-distance detection, the camera can cover 1-4 lanes, and the radar can cover 12 lanes at the same time, real-time analysis of non-human targets entering the field of view, accurate detection and distinction of pedestrians and Vehicles, and complete capture and recognition at the best distance, target capture rate > 99.9%, target feature recognition rate ≥ 98%, support 1-180 km/h speed measurement, ranging accuracy is ± 0.2m, speed accuracy is ± 0.2 m/s, the azimuth accuracy is ±0.3°.
②Deep fusion of radar and video data, real-time processingLeTV integrated design, built-in high-performance processor, real-time processing of video stream and millimeter-wave radar point cloud data, to achieve real-time fusion tracking processing of video targets and radar targets.
③ Integrate five functions into one to realize management upgradeThe equipment integrates the five functions of road video surveillance, traffic target recognition, traffic flow detection, traffic incident detection, and traffic information release, allowing traffic management to realize “insight into the overall situation” and truly form a visual traffic supervision without blind spots.
Its highly integrated design of “one machine with multiple functions” can assist the reasonable timing of traffic signals and improve the efficiency of intersection traffic; the holographic technology presents the real road world and realizes the digital twinning of traffic; “Wisdom Road” builds a solid foundation.
④Pragmatic innovation, engineering ease of use design
The equipment adopts a highly integrated design such as video camera, radar, warm light fill light, spirit level, double rainproof hat brim, etc., which is small and easy to use. The engineering structure is also very user-friendly, and its radar angle of view and lens angle of view have been preset according to the actual scene, and the construction is efficient and convenient.
Typical Application Scenarios
smart intersectionAccurate collection of traffic flow, assisted dynamic timing optimization of traffic signals and holographic presentation control, scientific allocation of road rights, shortened delay index by more than 20%, and improved traffic efficiency at intersections.
road blind spot
For blind spot scenes such as curves, up and down steep slopes, foggy mountainous areas, and tunnel entrances, the device can accurately detect the speed and distance of oncoming vehicles, and publish them to the opposite traffic guidance screen in real time, providing two-way early warning to prevent road traffic accidents .
Meet the requirements of vehicle-road coordination for roadside perception, realize road safety early warning, provide data support for vehicle-road coordination applications, provide assistance for artificial intelligence driving, and also provide more reliable environmental support for single-vehicle automatic driving technology.
Editor: Huang Fei