Embedded vision computing impacts how devices interpret and interact with their environment. By integrating advanced machine learning algorithms directly into edge devices, such as cameras, drones or industrial sensors, these systems can process visual data locally without relying on cloud connectivity. Advantages include reduced latency, enhanced privacy and real-time decision-making capabilities.
Embedded Vision
Embedded Vision
Unlock vision intelligence at the edge
Embedded Vision Applications
-
Industrial
Enabling machines to intelligently interpret visual data in real-time, enhancing accuracy, safety, productivity and automation.
-
Consumer
Empowers consumer devices to interpret visual data instantly, delivering intuitive experiences, enhanced safety, personalized interactions and smarter home automation.
-
Automotive
Enabling real-time perception and intelligent decisions that enhance vehicle safety, driver assistance, autonomous navigation and overall road experience.
Related Technologies
Related Products
Applications Processors
Microcontrollers
Automotive Processing
Secure connectivity
Security
Embedded Vision FAQs
Learn more about NXP’s embedded vision portfolio.
I’m not an expert in Machine Vision, where should I start?
NXP provides easy to use, industry-standard frameworks to enable rapid building of Vision applications, including connecting a camera, image preprocessing, machine learning inferencing with well-known models, and post-processing. If you are using a Linux system, see the i.MX Smart Fitness example on our Application Code Hub. If you are using an RTOS application, see the Multiple Face Detection on MCXN947 example.
Does NXP offer SoCs with integrated Neural Processing Units?
Yes. NXP SoCs that embed an integrated Neural Processing Unit range from low-cost, low-power 32-bit MCUs to powerful 64-bits Applications Processors. Please refer to nxp.com/ai to access the full list.
Does NXP offer SoCs with integrated Image Signal Processors?
Yes. Please see our i.MX 8M Plus and i.MX 95 SoCs, offering integrated ISPs, supporting multiple high resolution (up to 12MPix) cameras (up to 8).
I need to build a very simple and low-cost vision system. What would you recommend?
Look at our MCX N series MCUs, which integrates our eIQ® Neural Processing Unit and can process low-resolution video. If you need a bit more capability, our i.MX RT700 Crossover MCUs and i.MX 93 in the MPU/Linux space are good upgrades.
NXP Agrees to Acquire Edge AI Pioneer Kinara to Redefine the Intelligent Edge
Delivers high-performance neural network processing with advanced generative AI to create transformative edge use cases.
NXP Embedded Vision Partners
| Partner | Technology | Description |
|---|---|---|
| Arcturus | Solutions | Edge software and hardware technology provider, specialized in voice, video, vision, AI/ML |
| Au-Zone Technologies | Solutions | AI/ML software solutions for IoT edge devices, focusing on intelligent vision technology |
| Eta Compute | Tooling | No-code MLOps tool chain, covering everything from model development, deployment and management for the edge ML |
| Framos GmbH | Camera | Optical sensor modules for embedded vision |
| IMDT Ltd | Camera | AI and vision solutions and products for the edge |
| InnoWave Design LLC | Camera | End-to-end imaging solutions for AI-enabled vision solutions |
| iRider | Solutions | Safety software platforms for rider assistance, leveraging AI, vision and connectivity |
| nota | Tooling | Advanced AI models optimization tooling for any device and industry |
| TechNexion Ltd | Tooling | Embedded vision and computing blocks for the edge |
Design Resources
Additional Documents
Intellectual Property Aspects of ML
This white paper sets out to add clarity to the legal background and challenges for the IP aspects of ML for data collection and model building.
Read the white paperMachine Learning at the Edge
Insights into machine learning inference at the edge, such as uninterrupted processing, lower latency and user privacy.
Sign in to download ebookeIQ Gen-AI Flow
eIQs Gen-AI Flow provides building blocks like optimized Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) to make generative AI applications accessible on edge devices.
Read the white paperNXP Ethical Framework for Artificial Intelligence
See how we're teaching and embedding morals into machines through robust AI ethics.
Read the white paper