In varifocal and zoom models, Adaptive IR selectively activates wide-angle IR LEDs for short focal lengths and narrow-angle LEDs for longer distances, optimizing coverage and eliminating dark edge areas.

For PTRZ cameras with adjustable lens positions, independent external IR LED banks activate only those aligned with the lens direction, ensuring efficient long-range IR illumination.

In real-world environments, people and vehicles naturally move across entrances, corridors, buildings, and camera zones.

While cameras and analytics can capture key moments, those moments remain fragmented across different views and timelines.

Without a way to connect these fragments, operators must manually search footage camera by camera, making it difficult to reconstruct events or follow movement across a site.

Operators can select attribute-based references and receive a ranked list of related results across cameras and time periods, making cross-camera review and investigation far more efficient.

Most surveillance footage includes long stretches where important objects appear too small, unclear, or surrounded by unnecessary background.
Operators often rewind, replay, and manually scrub timelines just to locate a single clear frame.

When time matters, finding the right moment shouldn’t feel like searching for a needle in a haystack.

Alongside the image, cameras generate rich metadata describing:

• Object type (person/vehicle categories)
• Vehicle color
• Clothing colors (top/bottom)
• Additional attribute data for VMS search

These BestShots and metadata are sent to the VMS/NVR, enabling fast retrieval, filtering, and forensic review without manual frame-by-frame playback.

Surveillance cameras capture countless people and vehicles every day. But without intelligent classification, they remain just moving shapes, even when image quality is high.

Operators can see activity, yet struggle to determine who is involved, what type of vehicle passed by, or which attributes matter for investigation. Without object-level understanding, identifying suspects, tracing vehicles, or filtering events by appearance becomes slow, manual, and unreliable.

By combining real-time detection, precise classification, and metadata-based search, Object Detection & Classification transforms raw video into structured, actionable intelligence.

Many security incidents begin with sound — a scream, a sudden impact, or breaking glass. But these early warning signs often occur outside the camera’s field of view or in privacy-sensitive areas where cameras cannot be installed.

Without audio intelligence, these signals go unnoticed, delaying response and increasing risk.

All processing is performed through edge-based analytics, eliminating the need for a separate server while enabling real-time detection and alerts. For camera-restricted spaces, Audio Beacon (SPS-A100M) delivers the same intelligent audio detection, providing a powerful security alternative without violating privacy.

By integrating sound into surveillance, these solutions elevate security from a purely visual system to a comprehensive “see and hear” approach for a more complete understanding of every event.

Traditional surveillance captures massive amounts of footage but most of it remains unused. Operators are left scanning screens manually, reacting after incidents occur, and struggling to extract meaningful insights from raw video streams. Without intelligent analytics, security systems stay reactive instead of proactive, and operational data remains hidden in plain sight.

Crowd Counting analyzes people counting data per defined zone in real time to monitor crowd density and distribution.
The system accurately counts up to 500 people per zone, providing reliable insights into congestion levels and crowd buildup.
This enables operators to identify overcrowding risks early and take proactive measures for safety and flow control.

Queue Management monitors up to three defined queue zones simultaneously, displaying real-time people counting data directly on the web viewer.
The system triggers alarms based on the number of people per zone, enabling immediate response to congestion. It also supports event action settings, allowing automated system responses based on predefined queue conditions.

AI-based Line Crossing and IVA Area Detection monitor virtual lines and defined zones in real time. The system identifies people and vehicles, triggering alerts only when valid objects cross a line or enter a restricted area.
Unlike pixel-based detection, it ignores shadows, reflections, or random motion — ensuring precise intrusion and perimeter monitoring for facilities, borders, and restricted zones.

Social Distancing Detection monitors the proximity between multiple people simultaneously in real-time, identifying when predefined distance rules are violated. Once distances are calibrated through pre-setting, the system continuously measures spacing and triggers real time detection events when thresholds are breached. All processing is performed through edge-based AI analytics, eliminating the need for a separate server while ensuring fast, reliable response.

Surveillance often involves personal data — faces, license plates, identities. Static masks can’t follow motion, and manual setup increases risk of error or exposure. As privacy regulations grow stricter, operators need a way to protect individuals and still
preserve essential evidence when incidents occur.

As surveillance system expand, raw video data can quickly overwhelm bandwidth and storage capacity. Without optimization, even the best cameras risk lag, dropped frames, and excessive costs. Managing video shouldn’t mean compromising quality – and that’s where WiseStream steps in.

Here’s how WiseStream’s core mechanism works:

• AI Object Detection: An AI model precisely identifies user-defined key objects such
as people and vehicles, within the video stream.
• Differential Quality Application: The Region of Interest (ROI), where objects are
detected, is encoded at a high quality to preserve detail. In contrast, the Non-ROI is
subjected to a higher compression rate to reduce data volume.
• Efficient Static Area Compression: The system’s AI-based analysis minimizes
unnecessary data generation in static areas, ensuring efficient compression of the
entire scene.

Video surveillance doesn’t always happen in ideal conditions. Backlit entrances wash out faces. Dim corridors hide movement. Traditional noise reduction often sacrifices sharpness, while excessive WDR tuning creates unnatural contrast.

The result? Missed details, unreliable analytics, and wasted storage on unusable footage.

Conventional noise reduction techniques often blur details or leave ghosting effects. Our AI-NR takes it a step further — the AI model is meticulously trained on the unique noise characteristics of each specific image sensor, enabling it to distinguish real image data from noise at a pixel-by-pixel level.

Rather than simply smoothing over imperfections, it selectively removes noise while preserving fine textures, sharp edges, and natural colors, even in extremely low-light environments. By eliminating unnecessary noise data at the source, the AI-NR technology not only delivers remarkably clear and detailed video, but also reduces bitrate for more efficient storage and network performance.

Unlike conventional systems that mainly controlled high-speed shutters to reduce blur, AI-based Prefer Shutter technology uses the latest AI to identify object appearance and movement. This allows it to automatically readjust the shutter speed to an optimal level, effectively improving noise while minimizing motion blur.

In scenes without movement, the shutter slows down to ensure low noise and bright, clear images. Conversely, when significant object movement is detected, the shutter speeds up to reduce blur and deliver crisp visual clarity.

AI is no longer an optional enhancement — it is the foundation of modern video security.
With larger environments, higher resolutions, and an ever-growing volume of video data, operators need systems that can process information intelligently, reduce noise, surface what matters, and automate complex tasks.

Hanwha Vision’s AI brings clarity, speed, and insight to an industry where every second counts.