Automatic Vehicle Classification

PCC500-VFR(P)-V7


The Automatic Video Vehicle Classification System(AVVCS), mainly comprises a vehicle front camera, a vehicle rear camera, a panoramic camera of the vehicle body, a high-performance AI processor chip, and LED Light parts.

It integrates image acquisition, image processing, image capture, vehicle feature detection, vehicle classification, and license plate recognition, which can realize the automatic recognition and classification of the vehicle type. It has the characteristics of high accuracy of vehicle classification, easy installation, and easy maintenance. The AVVCS is ideal for use in the following scenarios:

Automatic card issuing system of highway toll stations, Fake plate vehicle investigation, ETC fee evasion management, Highways and customs weighing lanes, Vehicle classification of big-scale logistics parks and toll, container damage inspection, Container number identification, etc.


  • Vehicle Recognition Accuracy≥ 98%

  • Axle Recognition Accuracy≥ 99%

  • Wheel Recognition Accuracy≥ 98%

Application Scenarios

  • Highway Toll Stations

  • Customs

  • Wharf Container Tally

  • Large Logistics Park and Agricultural Market


Function

◉ Vehicle Classification

The device can automatically recognize motorcycles, 2-4 axle cars, small trailers, 2-9 axle trucks, and other models.

◉ Multi-dimensional Feature Obtains of Vehicles

The device can identify the feature information of the vehicle, such as axle number, wheelbase, axle type, and wheel number.

◉ Video Detection of Vehicles

The device can detect and automatically segment vehicles through real-time video streaming and independently complete vehicle detection and model analysis without grating and vehicle detector assistance.

◉ HD Image Output

The device can automatically capture HD images of the vehicle's front and rear and HD panoramic images of the vehicle's side body to investigate afterward.

◉ License Plate Recognition

The device can automatically recognize the vehicle license plate number.

◉ Character Overlay

Support for overlaying time, custom characters, etc. on H.264 video streams and resulting images.

◉ Fill Light Control

The device supports photosensitive control and automatically controls the fill light on and off according to the environment's brightness.

◉ Camera Self-adaption

The device can automatically control the camera parameters in an all-weather, self-adaptive external environment.

Features

◉ High Accuracy of Vehicle Classification

The vehicle classification accuracy is over 98%, much higher than the accuracy of the pressure sensing vehicle type identifier which is 90%, which can reduce the workload of manual modification of vehicle classification results.

◉ Sufficient Visualization Evidence

It can automatically collect HD images of the vehicle’s front, rear, and panoramic side body, making all the details clear when investigating.

◉ Easy Installation and Debugging, Easy Maintenance

Its embedded integrated equipment is used to recognize vehicle types through video detection. Installation, debugging, and maintenance are very easy. There is no need for a ground sensing coil and grating separator, no road surface damage, and no need to worry about the ground sensing coil being crushed by vehicles.

Specification

Front/Rear Camera Sensor Type 5 MP 1/2.8" Progressive CMOS
Effective Pixels 2592(H)×1952(V)
Shutter Auto/Manual, 7µs~40ms, step 7µs
Min. Illumination Color 0.012Lux@(F2.0,AGC ON)
Lens Type Fixed-focal, focal distance 6mm
Panoramic Body Camera Sensor Type 5 MP 1/2.8" Progressive CMOS
Effective Pixels Based on the actual image stitching of the vehicle
Shutter Auto/Manual, 3µs~20ms, step 3µs
Min. Illumination Color 0.024Lux@(F2.0,AGC ON)
Lens Type Fixed-focal, focal distance 1.6mm(fish-eye lens)
White Balance Auto/Manual
HLC Yes
WDR Digital WDR
Noise Reduction 2D NR; 3D NR
Video Compression H.264 High Profile、H.265 Main Profile
Video Bit Rate 2Kbps~16Mbps
Video Frame Rate 1fps ~ 25fps
Resolution 1080P(1920×1080)
Image Settings Automatic adjustment of Exposure Time, Exposure Control, Gain and White Balance
Video Stream 1080P(1920x1080)、720P(1280x720)、540P(960x540)、576P/D1(720x576)、576P/4CIF(704x576)、480P(720x480),the bit rate can be set
Vehicle Classification Accuracy ≥ 98%
Axle Number Recognition Accuracy ≥ 99%
Wheel Number Recognition Accuracy ≥ 98%
Output Contents 1 large image of the vehicle front, 1 large image of the vehicle rear, 1 panoramic stitching image of the vehicle body, no less than 5 seconds of video, additional information text, etc.
Network Port 2 100/1000M self-adaptive RJ45(WAN1/WAN2)
I/O Input 1 For external coil
I/O Output 1 Relay control signal output
RS-485 Serial Interface 1 RS-485
Network Protocol Supports multiple network protocols, including TCP/IP, HTTP, NTP, and RTSP, etc.
Access Standard ONVIF
Power Supply AC 110-220V, 50-60Hz
Power Consumption ≤ 160W
Mean Time Between Failure(MTBF) MTBF ≥ 30000 Hours
Mean Time To Repair(MTTR) MTTR ≤ 90 Seconds
Environment Temperature, -20℃ ~ +60℃ Air pressure, 86Kpa ~ 106Kpa, Humidity, 20% ~ 90% (no condensation)