Dahua DHI-NVR5432-16HP-AI/ANZ-PRO 32CH WizMind 16× PoE & Advanced AI Search Network Video Recorder

Key System Features

• 32 IP camera channels

• 16 built-in PoE ports (IEEE 802.3af/at) for simplified installation

• 4 SATA HDD bays (each up to 20 TB)

• Supports Smart H.265+, H.265, Smart H.264+, H.264, MJPEG formats

• Supports up to 32 MP camera resolution

• Dual VGA and Dual HDMI outputs (4K on HDMI)

• ONVIF and third-party camera support

• ePoE compatible (selected ports)

---

AI Engine & Analytics

AI Core Technologies

• Xinghan Large-Scale AI Models: Next-generation deep learning for enhanced scene understanding and analytics

• WizSeek: Natural-language video search based on text descriptions and semantic cues

• AcuPick: High-performance intelligent target retrieval across large video libraries

AI by Recorder

• Face detection and recognition

• Video metadata (people, motor vehicles, non-motor vehicles attributes)

• Perimeter protection (tripwire, intrusion, etc.)

• Smart Motion Detection (SMD)

• AcuPick (8-channel)

• WizSeek semantic search

• Text-defined alarm rules

AI by Camera (compatible cameras)

• Face detection/recoognition

• Perimeter protection

• SMD Plus

• Stereo analysis

• People counting & crowd distribution

• ANPR (license plate)

• Heat map

• PPE detection

• AcuPick

---

Performance Specifications

• Incoming bandwidth: up to 512 Mbps (AI disabled)

• Recording bandwidth: up to 512 Mbps

• Outgoing bandwidth: up to 384 Mbps

• Face database: up to 20 face databases with 200,000 images total

• Face detection (Recorder):

  • Xinghan enabled: 8 channels @ 12 faces/s each

  • Xinghan disabled: 16 channels @ 12 faces/s each

• Perimeter protection:

  • Xinghan enabled: 16 channels

  • Xinghan disabled: 32 channels

---

AI Search & Analytics Explained

AcuPick (Intelligent Target Search)

AcuPick indexes recorded video using deep learning so users can instantly retrieve clips with specific targets (people, vehicles, objects) without timeline scrubbing. It’s the backbone of fast forensic search functionality.

In practical terms: investigators spend less time watching footage and more time getting to the exact events they need.

Xinghan AI Model

Xinghan is Dahua’s large-scale AI framework that enables higher-order scene understanding. By analysing visual information with deep networks, Xinghan supports advanced analytics and semantic correlation across images.

It enables features beyond standard motion or perimeter analytics, providing richer context and interpretation.

WizSeek (Semantic Video Search)

WizSeek allows operators to search recorded video using descriptive language — e.g., search “man in red shirt near entrance after 5 pm.”
This goes beyond simple object detection, letting operators find relevant footage using human-friendly criteria, not just frame motion or metadata tags.

---

Comparison Table