Dahua IVD5148-3I is a 48CH 1U INTELLIGENT VIDEO DEVICE.
Main Processor: 64-bit high-performance multi-core processor
AI Card: 3, AI card mode is switchable
In Multimodal Large Model Engine: Supports all channels (16 targets/s) Wizseek with AcuPick camera
In Intelligent Analysis Engine: Supports up to 2 algorithms running simultaneously; Up to 8-ch 2 MP or 4-ch 4 MP or 2-ch 8 MP face recognition/IVS/Video Metadata/AcuPick by recorder; Up to 4-ch 2 MP or 4-ch 4 MP or 2-ch 8 MP PPE Detection/Work Behavior Analysis/People Behavior Analysis/Smart Object Detection by recorder
In Compression Engine: Up to 8-ch 2 MP/ 4-ch 4 MP/ 3-ch 5 MP/ 2-ch 8 MP/ 1-ch 12 MP intelligent video compression
AI by Camera: Face detection; face recognition; video metadata (human, vehicle, non-motor vehicle); IVS; license plate comparison; tripwire people counting; regional people counting; abnormal number of people in queue; smoking detection; call detection; face & body detection
AI by Recorder: Face detection; face recognition; video metadata (human, vehicle, non-motor vehicle); IVS
AcuPick: 1. AI by Recorder: Supports 48-channel 2 MP or 48-channel 4 MP resolution(16-channel for each AI card); 2. AI by Camera: Supports 32-channel 2 MP or 32-channel 4 MP resolution
Performance: AI by Recorder: 32-ch 2 MP or 32-ch 4 MP or 16-ch 8 MP WizSeek with AcuPick IP camera (Min. 1 AI card for WizSeek modeling); 32-ch 2 MP or 32-ch 4 MP or 16-ch 8 MP WizSeek with normal IP camera (2 AI cards for AcuPick by recorder, 1 AI card for WizSeek modeling); Each AI card for WizSeek modeling can analyze 16 targets/s
Search Language: English, Italian, Spanish, Russian, French, German, Portuguese, Polish, Czech, Dutch, Korean, Arabic, Turkish, Thai, Indonesian, Vietnamese, Traditional Chinese
Face Detection: 1. AI by Recorder: Supports 48-channel 2 MP or 48-channel 4 MP resolution(16-channel for each AI card); 2. AI by Camera: Supports 32-channel 2 MP or 32-channel 4 MP resolution; 3. Supports 6 properties: Gender, age, glasses, expression, face mask, and beard
Face Database Capacity: 1. 100 human face databases with 500,000 face pictures in total; 2. 5 passerby databases with 500,000 face pictures in total; The human face database and the passerby database share the total capacity and number of face images
Face Recognition: 1. AI by Recorder: 48-channel 2 MP or 48-channel 4 MP with normal IP camera(16-channel for each AI card); 32-channel 2 MP or 32-channel 4 MP with face detection IP camera (Totally 32 × 2 MP or 32 × 4 MP face pictures analyzed per second); 2. AI by Camera: 32-channel 2 MP or 32-channel 4 MP with face recognition IP camera; 3. Supports entries frequency alarm
Search by Image: Search by image in the face database, search by face or human body images, and 1:1 face recognition
License Plate Database Capacity: 1. 50 plate databases with 300,000 plate numbers in total; 2. Supports allowlist and blocklist
Vehicle License Plate Comparison: Supports 32-channel 2 MP or 32-channel 4 MP resolution with ITC
Video Diagnosis: Normal CPU Mode: 1. Supports 1-channel tour detection; 2. Supports 13 detection items: blurry video; low contrast; video tampering; video jitter; abrupt image change; image freeze; overexposure detection; video loss; noise detection; scene changing; static noise; interference fringe and color cast; Intelligent CPU Mode: Not supported
Intelligent Video Compression: 1. Compressed preview, playback and download of source stream on the platform; 2. Supports origin/compression dual stream live view and storage; 3. Supports compressing a variety of third-party code streams (ONVIF/RTSP); 4. Compression Resolution: 12 MP/8 MP/6 MP/5 MP/4 MP/3 MP/1080p/1.3 MP/720p/D1; 5. Each AI card supports up to 2-channel 12 MP/4-channel 8 MP/6-channel 5 MP/8-channel 4 MP/16-channel 2 MP and below
Industrial Algorithms: The industrial algorithm is an optional installment
Work Behavior Analysis: 1. AI by Recorder: Without large model (by default): 24-channel 2 MP/4 MP/8 MP (8-channel for each AI card); With large model (optional): 12-channel 2 MP/4 MP/8 MP (4-channel for each AI card); 2. Rules: AI by Recorder: calling, using mobile phone, sleeping while on duty, leave post detection, smoking and abnormal number of people
PPE Detection: 1. AI by Recorder: 24-channel 2 MP/4 MP/8 MP (8-channel for each AI card); 2. Rules: AI by Recorder: hat/safety helmet detection, Balaclava, workwear detection (detect by color attributes or registered database), face mask detection, goggles detection, seatbelt detection, safety vest detection, shoes detection, gloves detection, mask detection, exposed hair detection and arm band detection
Smart Object Detection: 1. AI by Recorder: 24-channel 2 MP/4 MP/8 MP (8-channel for each AI card); 2. Rules: AI by Recorder: smart missing object and smart abandoned object; 3. Object Type: non-motor vehicles, luggage, bags and boxes
People Behavior Analysis: 1. AI by Recorder: Without large model (by default): 24-channel 2 MP/4 MP/8 MP (8-channel for each AI card); With large model(optional): 12-channel 2 MP/4 MP/8 MP (4-channel for each AI card); 2. Rules: AI by Recorder: fighting detection, fall detection, running detection and tailing detection
Access Channel: 48-channel
Network Bandwidth: Incoming bandwidth: 256 Mbps Recording bandwidth: 256 Mbps Outgoing bandwidth: 256 Mbps
Buy Dahua IVD5148-3I in Dubai, UAE. Check Price & Availability with KOMPLEKT for the required configuration, current price and stock confirmation.