Skip to main content

Person Tracking

Follow individuals seamlessly across your entire camera network. Visylix Person Tracking uses appearance-based Re-ID to maintain consistent identities as people move between overlapping and non-overlapping fields of view.

About Person Tracking

Visylix Person Tracking fuses real-time object detection with a deep re-identification (Re-ID) embedding model that encodes each person's visual appearance into a compact descriptor. When a tracked individual exits one camera's view and enters another, the system matches descriptors across cameras to maintain a single continuous track. Trajectory analysis maps full movement paths through your facility, revealing dwell zones, transition patterns, and flow bottlenecks. The model handles appearance changes caused by lighting shifts, partial occlusion, and crowded scenes, delivering reliable cross-camera continuity in environments ranging from shopping malls and airports to corporate campuses and stadiums.

Key Features

Engineered for accuracy, speed, and seamless integration into enterprise workflows.

Cross-Camera Re-ID

Maintain a unified identity for each person as they move across multiple cameras, even when views do not overlap and minutes pass between sightings.

Appearance-Based Matching

A 256-dimension appearance embedding captures clothing color, texture, body shape, and gait characteristics for robust matching under varying conditions.

Trajectory Analysis

Reconstruct full travel paths within a facility, identifying entry points, exit points, dwell times, and the sequence of zones each person visits.

Zone Transition Mapping

Define custom zones on a floor plan and automatically record every transition between them, powering heatmap and flow-diagram visualizations.

Occlusion Handling

Predictive motion models and appearance re-matching recover tracks after brief occlusions caused by pillars, crowds, or other obstructions.

Real-Time Performance

Process up to 30 fps per camera with GPU-optimized inference, supporting simultaneous tracking across hundreds of concurrent streams.

Real-World Use Cases

Deployed across industries to deliver measurable value from day one.

Retail Analytics

Map shopper journeys from entrance to checkout. Understand aisle traffic, dwell times at displays, and the path-to-purchase for layout optimization and conversion improvement.

Security Surveillance

Track persons of interest as they move through a multi-camera network. Automatically generate alerts when subjects enter restricted zones or deviate from expected routes.

Crowd Management

Monitor people flow through choke points, corridors, and gathering areas. Detect congestion before it becomes dangerous and guide real-time crowd control decisions.

Smart Buildings

Understand occupancy patterns floor by floor and room by room. Feed data into HVAC, lighting, and elevator systems to optimize energy usage and occupant comfort.

Technical Specifications

Re-ID Accuracy (Rank-1)95.4% on Market-1501
Embedding Dimensions256-D feature vector
Matching Latency< 10 ms per query
Max Concurrent Tracks500+ per camera
Frame RateUp to 30 fps per stream
Camera HandoffOverlapping & non-overlapping
Track MemoryConfigurable (1 min to 24 hrs)
Supported ProtocolsRTSP, RTMP, HTTP-FLV, WebRTC
HardwareNVIDIA T4 / L4 / A100 / Jetson
DeploymentOn-premise, edge, or cloud

Ready to Get Started?

Discover how Visylix AI models can transform your video infrastructure into an intelligent, automated system.