T4 Vehicle 360 ° Dynamic Image Intelligent Detection System

T4 Vehicle 360 ° Dynamic Image Intelligent Detection System

The system mainly consists of a front-end acquisition module and a back-end control and operation unit. The vehicle 360 ° dynamic image intelligent detection system can achieve intelligent diagnosis, analysis, and prediction of visible components of subway vehicle appearance, dynamically detect faults in visible components such as roof, side, bottom, and current collection devices, and automatically alarm, realizing 360 ° all-round detection of trains. It is mainly used for online imaging of vehicle body storage and autonomous defect detection.

SHENGHAO

Full color image, high-definition details
Using multiple sets of acquisition units to perform 360 degree panoramic scanning of all visible components of the vehicle's exterior can clearly present 0.5mm detailed images, highly restore the scene, full-color images
Detecting anomalies and intelligent recognition
It can quickly identify common problems such as screws (loose, falling off), square hole locks (facing off), cables (falling off), box covers (falling off), valves (facing off), crack damage, grease leakage
One car, one image, accurate image recognition
By using train carriages as units and accurately matching information such as train and carriage numbers, the condition of the carriages is monitored comprehensively. The images support zooming in and out and quick passing
Unified platform, data fusion
The detection and alarm method based on component data tree can achieve data fusion with train detection robots, unified platform management, and a combination of coarse and fine detection

Main functions

Automatic identification of vehicle number
Direction of arrival and speed detection
Deformation, detachment, and damage detection of visible components of car body windows
Key components of the vehicle's bottom running gear and side bogie, including traction devices, motor covers, brake shoes, or brake discs
Detection of missing, deformed, and foreign objects in gearboxes, axle boxes, etc
Detection of missing and deformed bolts, connectors, brake components, etc. on the side and bottom of the vehicle's running gear
Detection of missing or deformed roof antennas, porcelain bottles, wiring, etc

Technical Parameters

Accuracy of vehicle number recognition Adapt to vehicle speed Outdoor equipment protection level 2D image accuracy 3D image accuracy: ≥99.99% 3-30km/h， Bidirectional detection ≥IP65 ≤1mm/pixel ≤±3mm

Full color image, high-definition details

Using multiple sets of acquisition units to perform 360 degree panoramic scanning of all visible components of the vehicle's exterior can clearly present 0.5mm detailed images, highly restore the scene, full-color images

Detecting anomalies and intelligent recognition

It can quickly identify common problems such as screws (loose, falling off), square hole locks (facing off), cables (falling off), box covers (falling off), valves (facing off), crack damage, grease leakage

One car, one image, accurate image recognition

By using train carriages as units and accurately matching information such as train and carriage numbers, the condition of the carriages is monitored comprehensively. The images support zooming in and out and quick passing

Unified platform, data fusion

The detection and alarm method based on component data tree can achieve data fusion with train detection robots, unified platform management, and a combination of coarse and fine detection