Auto-Switch Video Analysis2 engines in total,one engine can support up to 16ch,(switch time between 10 to 3600 seconds)
Live Video Analysis2 engines in total,one engine can support up to 4ch,(4ch@4MP，2ch@6MP，1ch @8MP)
Offline Picture Analysis4 image per GPU per second or 1 image per GPU per second ( Large model)
Scheduled Capture Analysis2 engines in total,one engine can support up to 16ch, (switch time between 3 to 3600 seconds)
Auxiliary Interface
Alarm In/Out16/9
CapacityUp to 16 TB capacity for each HDD
Ctrl 12VControllable 12 VDC, 1 A power output for external alarm device; The power will be turned on when the alarm output is triggered.
*: The Ctrl 12V power is controlled by alarm output 9.
DC 12V12 VDC, 1 A power output
SATA8 SATA interfaces; 3.5-inch HDD
Serial Interface1 RS-232, 1 RS-485 (full-duplex)
USB InterfaceFront panel: 2 × USB 2.0; Rear panel: 2 × USB 3.0
eSATA1 eSATA interface
Decoding
Decoding CapabilityUp to 2-ch@32 MP/2-ch@24 MP/4-ch@12 MP/8-ch@8 MP/16-ch@4 MP/32-ch@1080P
Decoding FormatH.265/H.265+/H.264/H.264+
Recording Resolution32 MP/24 MP/12 MP/8 MP/6 MP/5 MP/4 MP/3 MP/1080p/UXGA/720p/VGA /4CIF/DCIF/ 2CIF/CIF/QCIF
*: After ultra HD resolution mode is enabled, the NVR supports up to 8-ch 32 MP/24 MP IP video inputs.
AI by NVRPre-installed AI Algorithm Package
Human (3 Algorithm models )
1. Facial recognition:
2. People counting: regional people counting, people counting statistics
3. PPE: safety helmets and reflective vests detection
Vehicle (2 Algorithm models )
1. Vehicle attributes: vehicle brand,vehicle color, vehicle type
2. Non-motorized vehicle detection: motorcycle, bicycles
Dangerous Event (2 Algorithm models )
1. Fire and smoke detection
2. Occupation of fire escape
Behavior (3 Algorithm models )
1. Street level behavior : falling down detection, people gathering detection, fast
moving detection, physical conflict detection
2. Indoor: playing phone, sleeping on duty, absence, number of people allowed, overstaying detection
3. Perimeter: intrusion detection, line crossing detection, region entrance detection,
region exiting detection
4. Smoke detection, call detection
AIOP (Up to 16 models)
Model type: detection model, classification model, mixed model (detection model and
classification model)
1 engines in total, One engine can support four algorithm models;
each algorithm model supports one real-time video stream analysis or a maximum of 8
auto-switch video analyses (switch time between 10 to 3600 seconds)
Engine3 * KT2 (2 for live video analysis, 1 for lager model algorithm secondary analysis)
Network
APIONVIF (profile S/G); SDK; ISAPI
Compatible BrowserIE11, Chrome V57, Firefox V52, Safari V12, Edge V89, or above version
