NVIDIA® Jetson Orin™ Reference Design

NVIDIA® Jetson Orin™ reference designs are built for embedded vision systems that need reliable camera integration, local image processing and scalable development. The Imaging Source platform combines embedded cameras, carrier board options and software support for MIPI CSI-2 camera, GMSL2 camera and FPD-Link III camera integration with NVIDIA Jetson Orin. This platform is suitable for AI vision, robotics, quality inspection, logistics, smart city systems, agriculture and OEM machine vision systems based on NVIDIA Jetson Orin.

NVIDIA® Jetson Orin™ Reference Design collection

NVIDIA Jetson Orin platform for embedded vision

The NVIDIA Jetson Orin platform is designed for embedded vision applications that need camera input, image processing and AI performance in one compact setup.

It gives engineering teams a validated starting point for evaluating cameras, testing software and building application-specific vision systems with NVIDIA Jetson Orin.

Why choose NVIDIA Jetson Orin for embedded vision?

The NVIDIA Jetson Orin platform is useful when a system needs multi-camera support, strong processing performance and a clear development path.

01

MIPI CSI-2, GMSL2 and FPD-Link III cameras

Support for multiple embedded camera interfaces helps engineers choose the right camera connection for the system layout.

02

Up to 4 cameras per system

Multi-camera support enables several viewpoints for robotics, inspection, logistics and AI vision applications.

03

0.4 MP to 20 MP camera resolutions

Camera options can support compact sensing, high-resolution inspection and application-specific image acquisition.

04

Sony and onsemi CMOS sensors

The platform supports camera options based on established CMOS sensor technologies for embedded vision development.

05

Jetson Orin NX and Orin Nano options

Different SoM options help match processing performance, RAM and system requirements to the application.

06

BSP and JetPack 6.x support

Board support package and JetPack support help reduce setup effort during embedded vision development.

07

Camera driver and IC Imaging Control 4 SDK

Driver and SDK support helps engineers move from first image acquisition to a working camera setup faster.

08

GStreamer, OpenCV and MVTec HALCON

Framework compatibility supports image acquisition, image processing and machine vision development workflows.

NVIDIA Jetson Orin reference design for camera integration

The NVIDIA Jetson Orin reference design helps simplify embedded vision camera integration by combining the main hardware and software elements in one platform.

Instead of selecting each component separately, engineers can start with a validated reference design and adapt it for production with NVIDIA Jetson Orin.

NVIDIA Jetson Orin reference design for camera integration

NVIDIA Jetson Orin camera interface options

The NVIDIA Jetson Orin platform supports multiple embedded camera interfaces for different cable length, placement and system layout requirements.


MIPI CSI-2 cameras can connect directly, while GMSL2 and FPD-Link III cameras require a deserializer adapter board for NVIDIA Jetson Orin.

NVIDIA Jetson Orin with MIPI CSI-2 cameras

MIPI CSI-2 cameras are suitable for compact NVIDIA Jetson Orin systems where direct, low-latency camera integration is required.

Show MIPI CSI-2 Cameras
NVIDIA Jetson Orin with GMSL2 Cameras

GMSL2 cameras are suitable for NVIDIA Jetson Orin systems that need longer cable distances, robust transmission and remote camera placement.

Show GMSL 2 Cameras
NVIDIA Jetson Orin with FPD-Link III cameras

FPD-Link III cameras are suitable for NVIDIA Jetson Orin systems that require reliable remote camera integration with flexible camera placement.

Show FPD-Link III Cameras

NVIDIA Jetson Orin multi-camera support

The NVIDIA Jetson Orin reference design supports multi-camera embedded vision systems for applications with several camera positions or viewpoints.

Depending on the configuration, the platform can support up to 2 MIPI CSI-2 cameras or up to 4 GMSL2 or FPD-Link III cameras with NVIDIA Jetson Orin.

NVIDIA Jetson Orin hardware configurations

The NVIDIA Jetson Orin platform is available with NVIDIA® Jetson Orin™ NX and NVIDIA® Jetson Orin Nano™ system-on-module options.

Different configurations support different RAM, AI performance and camera count requirements for embedded vision applications using NVIDIA Jetson Orin.

NVIDIA Jetson Orin reference design components

The NVIDIA Jetson Orin platform includes software support for camera integration, system setup and embedded vision development.


Support can include a board support package, JetPack 6.x, camera drivers and IC Imaging Control 4 SDK for development with NVIDIA Jetson Orin.

NVIDIA Jetson Orin Software Support for Embedded Vision

The NVIDIA Jetson Orin platform is supported by software tools and development resources for camera integration, system setup and embedded vision development. Support can include a board support package, JetPack 6.x, camera drivers and IC Imaging Control 4 SDK for development with NVIDIA Jetson Orin.

Framework Compatibility and AI Vision Workflows

NVIDIA Jetson Orin supports common image processing and machine vision frameworks such as GStreamer, OpenCV and MVTec HALCON. These frameworks help engineering teams build image acquisition, processing and AI vision workflows for applications such as robotics, inspection, classification, object detection and autonomous systems.

NVIDIA Jetson Orin applications

The NVIDIA Jetson Orin platform is suitable for embedded and industrial vision systems that need camera integration and local processing.

Embedded AI vision

Local AI processing for camera-based detection, classification and analysis.

Robotics and autonomous systems

Camera integration for mobile machines, robots and autonomous platforms.

Quality inspection

Vision-based inspection for industrial processes, products and production lines.

Logistics and tracking systems

Embedded vision for identification, monitoring and process visibility.

Smart city and IoT vision

Camera systems for connected infrastructure, monitoring and edge applications.

Agricultural monitoring and automation

Vision integration for outdoor machines, crop monitoring and automation systems.

OEM embedded camera systems

Reference design support for product-specific embedded vision development.

Multi-camera machine vision

Multiple camera viewpoints for inspection, positioning and scene analysis.

NVIDIA Jetson Orin from evaluation to production

The NVIDIA Jetson Orin reference design supports a clear path from early evaluation to production-ready embedded vision systems.


Validated camera hardware, carrier board design, software support and technical expertise help reduce integration risk when developing with NVIDIA Jetson Orin.

Need help choosing an NVIDIA Jetson Orin setup?

Selecting the right NVIDIA Jetson Orin setup depends on camera interface, camera count, resolution, frame rate, software and production requirements.

The Imaging Source can help compare carrier board configurations, cameras, software stacks and accessories for NVIDIA Jetson Orin.

Contact The Imaging Source for technical advice and support with your embedded vision setup for NVIDIA Jetson Orin.

Frequently asked questions about NVIDIA Jetson Orin

NVIDIA Jetson Orin is suitable when your embedded vision system needs local image processing, AI performance, camera integration and multi-camera support in a compact setup.

Use MIPI CSI-2 when the camera and processing platform can be placed close together. Use GMSL2 or FPD-Link III when longer cable distances or remote camera placement are required with NVIDIA Jetson Orin.

NVIDIA Jetson Orin reference designs can support up to 4 cameras per system, depending on the selected camera interface, carrier board and system configuration.

Yes. GMSL2 and FPD-Link III cameras require a deserializer adapter board when they are used with NVIDIA Jetson Orin.

NVIDIA Jetson Orin camera integration can be supported with a board support package, JetPack 6.x, camera drivers and IC Imaging Control 4 SDK, depending on the selected setup.

NVIDIA Jetson Orin supports frameworks such as GStreamer, OpenCV and MVTec HALCON for image acquisition, processing and embedded AI vision workflows.

To select the right NVIDIA Jetson Orin setup, define the camera interface, camera count, resolution, frame rate, cable distance, software framework and production requirements.