Edge-based Pedestrian AI Safety System
1. Edge AI Camera Traffic Detection and Perception Functions
• Vehicle Classification: Truck, Bus, Sedan, Motorcycles, Pedestrians, and Bicycles
• Vehicle detection within 50 meters of the intersection, including speed estimation, braking distance, and red-light violation prediction
• Pedestrian coordinate projection
• Pedestrian zone occupancy detection
• Intersection incident detection
• Directional pedestrian crossing count
2. Edge Computing Device Hardware Specifications
• CPU: NVIDIA Jetson Xavier NX
• Memory: 8GB LPDDR4X
• Operating temperature: -15°C to 60°C
• IO Rear: Dual GBE (with 802.3AT PSE) & Single GBE
USB 3.0 x4 & Micro USB x1
DB-9 x1 for RS-232/485 & CANbus
DB-15 x1 for 13-channel Digital I/O (programmable)
HDMI 2.0 output x1, HDMI 1.4 input x1
Antenna ports x4
• Dimensions: 181(W) x 136(D) x 62(H) mm
3. Camera Detection Unit
• 2MP CMOS image sensor
• 30 fps @ 1080p Full HD
• Motorized zoom lenses: 3.3–10 mm, 6–22 mm, or 5–50 mm; PL722F: 8 mm fixed focal length (optional 6 mm), fixed aperture F1.8 lens
• Built-in digital WDR enhancement
• Integrated infrared illuminator with effective range up to 15 meters
Inquiry