AI Vision — Object Detection & Tracking
Targeting and tracking
Inogate's AI Vision system is a compact and powerful computer vision solution built for real-time object detection, classification, and tracking in demanding operational environments. Trained on an 80,000-example dataset, the system achieves an average confidence score of 0.88 and identifies objects as small as 16×16 pixels even under challenging conditions.
The system supports multiple resolutions — delivering up to 64 FPS at 630×480 and 37 FPS at 1920×1080 — and is compatible with both standard and infrared sensors. It is provided as a C++ SDK for seamless integration into diverse platforms, with custom OEM solutions available for nonstandard hardware configurations.
Applications: area surveillance, security monitoring, UAV/UGV integration, manned vehicle systems.
The AI Vision system is compatible with the following single-board computers, enabling flexible deployment across a wide range of platforms:
Raspberry Pi 4 — A high-performance single-board computer featuring a quad-core ARM Cortex-A72 processor and up to 8GB RAM, ideal for demanding real-time vision processing tasks.
Raspberry Pi Zero 2W — An ultra-compact and lightweight board powered by a quad-core ARM Cortex-A53 processor, perfect for weight-sensitive UAV integrations where minimal footprint is essential.
Orange Pi 5 — A powerful board equipped with a Rockchip RK3588S octa-core processor and a dedicated NPU, delivering high-speed AI inference for complex detection and tracking scenarios.
Orange Pi Zero 2W — A compact and energy-efficient board offering a solid balance between performance and size, suitable for embedded and resource-constrained deployments.
