Supports mainstream cameras of ONVIF and RTSP protocols
H.264/.H265/Smart H.264+/Smart H.265+/MJPEG, H.265 auto switch
1 VGA/1 HDMI simultaneous video output by default, the maximum resolution of HDMI is 4K; and non-simultaneous video output is also supported, the maximum resolution is 1080P
1-channel face detection and recognition (AI by NVR); or 2-channel perimeter detection (AI by NVR); or 4-channel SMD (AI by NVR); up to
10 face databases and 5,000 face images
All-channel AI by camera: face detection and comparison, perimeter protection and SMD
This AI network video recorder is a highly cost-effective economical product, providing functions such as local live view, multi-screen display, real-time storage on local HDD, fast operation by mouse, and remote management and control. This product is suitable for face detection and recognition at entrances/exits, and perimeter protection of people and vehicles.
Functions
New User Interface
Local and web interfaces adopt the new UI 4.0 style which caters to users habits. Functional modules are divided more clearly and easier to understand.
Decoding (Smart H.265+/Smart H.264+)
Dahua smart encoding (H.265+/ H.264+) is the optimized implementation of (H.265/H.264) codec that uses a scene-adaptive code rate control algorithm, providing more efficiency in the encoding process and higher video quality, and reducing the cost of storage and transmission significantly. This NVR supports the corresponding decoding.
SMD Plus
SMD Plus, referred to as Smart Motion Detection Plus, which is an upgrading version of SMD that greatly improves the alarm accuracy by loading deep-learning algorithm. It analyzes person and vehicle shapes based on motion detection and sends alarms only when person or vehicle intrudes.
Face Detection
Face detection is to detect if there is any human face appearing in the video. This technology adopts a deep learning algorithm to support face detection, tracking, optimization and capturing, and then output the best face snapshot.
Face Recognition
Face recognition is to figure out the captured face with target features, and compare them with the face database to identify the personnel.
Perimeter Protection
Based on the deep learning algorithm, perimeter protection provides accurate human and vehicle classification detection. It grealy reduces false alarms by smartly detecting behaviors such as tripwire and intrusion according to the detected object in restricted areas (e.g., for human or vehicle) .
Technical Specification
System
Main Processor : Industrial-grade embedded processor
Operating System : Embedded Linux
Operation Interface : Web/operating in local GUI
Video
Access Channel : 4 channels
Network Bandwidth : 80 Mbps for access, 80 Mbps for storage and 60 Mbpsfor forwarding.
Multi-channel Playback : Up to 4 channels playback
Record Mode : Priority of record mode: Manual recording > alarm recording > motion detection recording > timed recording
Storage : Local HDD and network
Backup : USB device
Playback Function : 1. Play, pause, stop, fast forward, fast backward, rewind,, play by frame, 2. Full screen, backup (cut/file), partially enlarge, audio on/off
SMD Plus
Performance : 4 channels
AI Search : Search by target classification (Human, Vehicle)
Face Recognition
Performance : Max 8 face pictures/sec processing, 1 channel video stream face recognition
Stranger Mode : Detect strangers' faces (not in device's face database). Similarity threshold can be set manually.
AI Search : Up to 8 target face images search at same time, similarity threshold can be set for each target face image.
Database Management : Up to 10 face databases with 5,000 face images in total. Name, gender, birthday, address, certificate type, cerificate No., countries®ions and state can be added to each face picture.
Database Application : Each database can be applied to video channels independently.