Tunnel detection types

1. Accurate accident detection possible even in harsh environments (light scattering, airborne dust, weather effects, etc.)

2. Consistent video processing regardless of the number or size of objects
3. Completed price competitiveness through software and hardware optimization
4. Additional object recognition and exception handling possible for motorcycles/bicycles/horses etc., according to requirements.

• Collision! Collision accident detection
• REverse! Wrong-way detection
• fire! Fire detection
• PEOPlE! Pedestrian detection

Automatically recognizes moving and target objects within the CCTV surveillance range, and can automatically detect and alert to emergencies.
Able to detect accidents, stopping, wrong-way driving, fire, pedestrians, etc. (Ministry of Land, Infrastructure and Transport management guidelines)

- Significantly increased detection distance for emergencies (from the usual 80-100m, VisionON T over 200m)
- Able to distinguish between stopping due to congestion/accident
- ALT (Control Operator Led Automatic Learning Training) system established
- Continuous learning can improve detection performance (False operation rate reducible by 99%)

System configuration diagram

1. Separate system construction other than the existing VMS

2. possible to operate with existing VMS

Actual operation status

1. Continuous learning and update support reflecting site conditions (free for 1 year)

2. On-site calibration of the deep learning model (3 months) service provided
3. Free performance update service for 9 months post-installation
4. Total 1-year free performance check and update service (Calibration 3 months + 9 months)

• A Tunnel
• B Tunnel
• C Tunnel
• D Tunnel

Patents·Certificates

• Tunnel emergency
  detection system patent
• Software quality certification
  GS certification grade 1 acquired