Smart Highway Toll Collection Solution

Hardware Specification
Function Modules
Hardware Specification

SoC:

Qualcomm QCS6490 learn more
CPU1×A78@2.7GHz 3×A78@2.4GHz 4×A55@1.9GHz
AI12 INT8 TOPS
GPUAdreno 643 @812MHz
WiFi802.11ax, 2.4G/5G DBS, 2*2 MIMO
BT5.2
Encoder4K @30 fps
Decoder4K @60 fps

Memory & Storage:

RAM8GB
ROM128GB

Mainboard interfaces required:

USB-A3×2.0
USB-C1×3.0, with DP
HDMI_OUT×1
LAN1×Gigabit Ethernet
RS485×1

Full specification of the reference main board, pls check in <Get>

Function Modules
/api/v1/files/solution/detail_software_module_imgs/20250327030638_卓视4.png
  • Vehicle Detection
    Real-time vehicle passing recognition, vehicle capture accuracy rate ≥ 99.5 %.
  • Vehicle Separation
    Based on multi-dimensional vision and auxiliary sensor fusion analysis technology, high-precision automatic separation of vehicles is realized without other equipment. The separation accuracy rate ≥ 99 %.
  • License Plate Recognition
    It can recognize the characters, letters, numbers and colors of civilian vehicle license plates that meet the standards of "GA36-1992", "GA36-2001", "GA36-2007", new energy vehicle plates, emergency rescue special number plates, embassy plates, and consulate plates. In addition, depending on the regulatory requirements of different countries, the device can recognize license plates from more than 40 countries.
  • Image Output
    Based on intelligent 3D video scanning and image analysis technology, the system outputs complete front and rear photos, splices and restores panoramic high-definition photos of the vehicle body, license plate photos and a complete video of the car passing by for at least 5 seconds.
  • Vehicle and Axle Type Recognition
    Compliant with the JT/T 489-2019 Classification of Toll Vehicle Types, the system can accurately recognize toll vehicle types, axle types, axle numbers, wheel numbers, and other information. Recognition accuracy rate ≥ 99 %.
  • Vehicle type recognition
    Based on deep learning technology and massive data accumulation, it can recognize more than 30 types of vehicle segmentation, including but not limited to box trucks, articulated axle trains, cement pump trucks, trailers, station wagons, etc. Recognition accuracy ≥ 99%
  • Vehicle Profile Measurement
    Based on intelligent multi-dimensional sensors for vehicle scanning, the system can accurately recognize the length, width, and height of the vehicle and output data through vehicle feature recognition.
  • Passenger vehicle identification
    By identifying the "number of passengers" labeled on the vehicle body, it is possible to accurately determine the type of bus. Recognition accuracy ≥ 98.5%
  • Hazardous Chemical Vehicles Recognition
    Based on the captured images of the front camera and the detection and analysis of vehicle body features, hazardous chemi- cal vehicle feature information can be output.