Camera hardware

DC Vision Edge AI Camera

DC Vision Edge AI Camera

Integrated embedded platform for real-time vision, 3D perception and AI.

Combining synchronized global-shutter cameras, NVIDIA Jetson Orin NX processing and the DC Vision SDK in a compact industrial system.

 

DC Vision Systems develops camera systems for real-time embedded applications, designed as part of complete perception and data processing solutions.

The focus is not only on image acquisition, but on extracting relevant information directly at the source through integrated processing and optimized data pipelines.

Processing & AI

  • NVIDIA Jetson Orin NX
  • 8 GB or 16 GB RAM
  • 256 GB NVMe SSD
  • 2 × M.2 PCIe slots
  • Optional M.2 AI accelerator
  • DC Vision SDK for real-time vision and AI

Camera & Sensors

  • 2 × Sony IMX296 global-shutter image sensors
  • Synchronized stereo image acquisition
  • Interchangeable M12 lenses
  • Integrated IMU

Interfaces

  • M12 Gigabit Ethernet
  • USB-C – USB 3.2 Gen 2, up to 10 Gbit/s in host mode
  • M8 power and GPIO
  • Additional application-specific I/O via GPIO

Power & Environmental

  • 6–20 V DC input
  • Approx. 20 W power consumption, depending on workload and configuration
  • Ambient operating temperature: –20 °C to +65 °C
  • Housing designed for dust and water protection

Key Features

  • Compact stereo camera system with global shutter sensors, Full HD resolution and configurable frame rates
  • Integrated embedded processing unit (≤15 W power consumption)
  • Robust, water- and dust-resistant enclosure for indoor and outdoor use (IP65)
  • Multiple interfaces: USB, Ethernet, GPIO
  • Wireless connectivity: 5G/LTE and Wi-Fi for remote access and IoT integration
  • Real-time SDK execution without cloud dependency
  • Cost-efficient design based on standard components and manufacturability
  • High flexibility and customization: interchangeable image sensors, optics, processing units and interfaces
  • Scalable for medium to high-volume production
  • Extendable to a central processing unit: optionally configurable as a central controller with external, wired sensor modules

System-Oriented Design

Our camera systems are tightly integrated with software frameworks and processing pipelines, enabling efficient end-to-end solutions from sensor to application.

Instead of delivering raw image data, the systems generate structured and usable information for downstream processing.

Typical Outputs

  • 3D environment representation
  • Object positions, motion and trajectories
  • Semantic scene information
  • Confidence and quality metrics

Architecture Overview

The systems integrate image sensors, embedded processing and optimized data pipelines to enable low-latency operation and efficient data handling.

Processing is performed directly at the edge, reducing bandwidth requirements and enabling real-time decision-making.

Positioning

The camera systems are part of a broader embedded AI and processing architecture, enabling real-time perception directly on edge devices.