With the deployment of 5G networks and intelligent traffic vehicle-road collaborative systems across the country, cutting-edge technical solutions that enhance traffic safety are increasingly valued.

In the field of traffic signal control, traditional perception technologies such as geomagnetic, inductive loops, or cameras are only capable of counting the number of vehicles passing by, which cannot meet the demands of digitalized management for refined and intelligent control.

The Thunder Vision Fusion Integrated Machine has the attributes of all-weather and long-distance detection. In addition to counting traffic volume, it can also monitor and track the motion behavior characteristics of vehicles, and lock vehicle information. At the same time, the Thunder Vision Fusion Integrated Machine, combined with ultra-long-range perception, expands the detection range from dozens of meters to hundreds of meters behind the lane, providing significant support for refined management such as traffic congestion and pedestrian crossing warnings.

The Thunder Vision Fusion Integrated Machine fuses real-time video and radar data to output target vector data, achieving real-time vectorization and tracking of global targets.

Rockchip provides a chip solution RK3568 that can be applied to this device. Qiyang, based on the Rockchip processor, has developed a board product. The Qiyang RK3568 core board and development board are based on the ARM Cortex-A55 architecture quad-core processor, integrating the Mali-G52 GPU, built-in NPU, with rich extension interfaces, providing reliable hardware support for decision-making, control, and interaction of the Thunder Vision Fusion Integrated Machine.

The core board is based on the quad-core Cortex-A55 architecture with a clock speed of up to 2.0GHz, integrating the Mali-G52 GPU, providing strong image processing capabilities, which is helpful for real-time processing of traffic monitoring data and image recognition data; built-in independent NPU with a computing power of 1.0TOPS, capable of quickly identifying vehicles, monitoring traffic violations, and early warning of traffic accidents; at the same time, through data processing and analysis of traffic flow, it provides real-time and accurate data support for traffic management departments to improve the efficiency and level of traffic management.

Built-in dual 8M ISP, supporting simultaneous input of two cameras to achieve image stitching of blind spots, providing more complete image information; at the same time, through image analysis algorithms, it can achieve multi-target detection and tracking, improving the automation and intelligence of the traffic monitoring system.

Supports single 4K or multiple 1080P@30fps video decoding, supports simultaneous encoding and decoding as well as dynamic adjustment of bit rate, frame rate, and resolution to ensure the clarity of the monitoring screen and the synchronization of the monitoring screen, avoiding information delay or misplacement, achieving comprehensive monitoring and scheduling of traffic flow and vehicle status.

The core board supports up to 8GB of memory, full-link ECC, high-reliability design can effectively improve system stability, reduce the error rate in data transmission and storage, and ensure the integrity and security of data.

Supports 4G/5G, WiFi, providing stable network connections to ensure the stability of data transmission of video data, sensor data, and storage data; with rich interfaces such as MIPI, CAN, USB, RS232, GPIO, etc., it can access the original video stream and radar data stream; at the same time, it is convenient for peripheral expansion, and can realize linkage screens, alarms, and loudspeaker warning lights for speed reminder functions.

Optional industrial-grade configuration, industrial-grade temperature range design, with high stability, can meet the device's stable operation outdoors all day long.