Visual AI and sensor fusion software for enhancing camera-based systems.

Accurate, continuous visual monitoring is a vital cornerstone of industries that use camera technology to anticipate and prevent threats.

As the amount of sensor data continues to increase exponentially, so does the already-immense burden on the security providers tasked with keeping us safe.

Now more than ever, operators and maintainers demand systems that are easier to use and perform as promised. To achieve this, they need automation.

Vizgard ai for defence and security video wall


black fortigai logo with shield white background

Advanced platform-agnostic software combining visual Artificial Intelligence, sensor fusion and camera platform control to increase time-critical system autonomy.

vizgard drone ai


Reduce fatigue of false alarms with detection, classification and tracking of drones, birds, planes and helicopters

Defence & Security

See further than ever with image optimistion and autonomously verify PTZ camera detections
computer vision search and rescue

Search & Rescue

Airborne visual detection and identification of missing persons over land and sea


Trigger alerts for potential poachers and not the animals you seek to protect
computer vision drone delivery

Delivery & Survey

Safe, GPS-free and human-aware drone landing, and AI-assisted surveys and inspections

Maritime ISR

Find those who do not wish to be found with small object detection impossible to match by the human eye
Vizgard Partner Logos


Your Counter-Drone Solution

Avoid false positives from other sensors with automatic visual confirmation and accurate tracking of genuine unauthorised drones.

How it works

Connect your platform’s video streams to our cloud, edge or on-premises FortifAI application to gain actionable intelligence that dives far deeper than simple, YOLO-based object detection.

This information can be integrated with and displayed by your existing software, and fused with external sensor data wherever available, enabling FORTIFAI to reduce operator burden and increase system efficiency across a host of task automation needs.

vizgard overview diagram

Next-Gen Object Tracking

Our state-of-the-art single object tracker builds on the progressive academic research into deep neural networks to achieve superior tracking – confidently handling agile or never-before-seen targets that may leave the FOV.

Key Features


High performance and low-SWAP

High Accuracy

AI trained on our proprietary dataset of millions of real-world images


40ms latency on local 4k video, and sub-second via the internet

Track Anything

Mark a box around any object to begin state-of-the-art AI-powered single object tracking

Secure Encryption

DTLS and AES128-GCM encryption

Image Enhancement

When every second counts, impaired performance caused by external factors including heat turbulence, fog and platform movements can result in undetected threats or delayed rescue.

FortifAI allows operators to reduce image degradation without ever compromising on detection accuracy and tracking capability.


Streamline your integration of FortifAI with our plug-and-play VizEDGE computers.

By combining market-leading NVIDIA Xavier AGX and NX production modules with a selection of tested, approved carrier boards, these low-SWAP devices deliver unrivalled AI performance and an adaptable combination of available inputs and outputs. Enclosure optional.

Latest News


Your Unmanned System

Ensure you never overlook a thing with onboard AI that increases automation and safety.


Vizgard competitor comparison


The standard FORTIFAI detection model locates and classifies people, cars, drones, birds, planes, faces, helicopters and boats. Users can initiate single object tracking with a single click or let the system choose the highest priority threat. Need an object type that we don’t yet support? Any object can be tracked by simply drawing a box around a target, or we can train a custom detection model to satisfy your unique use case.

Although we detect faces as part of our face redaction offering, we do not apply any recognition algorithms that would tell us personal details about that person such as their name. We focus on detecting the presence of a person and their present activity, rather than who they are or where they come from.

We currently support RTSP, WebRTC, HTTP, MP4 files, YouTube, and SDI video inputs and can output RTSP and WebRTC. However, we are always increasing our supported formats so feel free to get in touch if you require a different format. Our VizEDGE AGX supports 2x Ethernet, 2x HD-SDI and 2x Composite video.

For edge applications, we recommend you use our VizEDGE AGX or VIZEDGE AIR platforms which have been optimised to give you the very best performance. However, you can also source your own edge hardware, provided it uses NVIDIA Jetson Xavier NX or AGX. For those who wish to use a Windows 10-based system, we recommend an i7 processor with 16GB RAM and Nvidia RTX 2070 graphics. Minimum GPU requirement GTX 1050.


Your Privacy

Protect the privacy of the public and your colleagues with real-time face redaction included as standard.

Contact us