Earn a Michigan Engineering AI Certificate — Stay Ahead of the AI Revolution
PowerBI Data Analyst - Create visualizations and dashboards from scratch
Overview
AI, Data Science & Cloud Certificates from Google, IBM & Meta — 40% Off
One plan covers every Professional Certificate on Coursera. 40% off Coursera Plus Annual.
Unlock All Certificates
Explore the evolution of Video4Zephyr (v4z) APIs from their foundational 2019 introduction to their transformation into a comprehensive video subsystem for the Zephyr real-time operating system in this 38-minute conference talk. Discover how the original basic APIs, while providing essential video device interaction capabilities, have revealed limitations as video hardware complexity and camera-based application demands have increased over the years. Learn about the current status of v4z and examine recent improvements that address these challenges, including the implementation of devicetree video port/endpoint mechanisms for better hardware abstraction. Understand the development of a unified video control framework that standardizes video device management and explore enhanced buffer management systems designed to optimize video data handling. Investigate support for memory-to-memory (m2m) devices that enable advanced video processing capabilities within the Zephyr ecosystem. Gain insights into how these architectural improvements shift common functionality and logic into the subsystem itself, significantly reducing the development burden on driver creators while advancing Zephyr's video stack toward a mature multimedia framework comparable to established video processing systems.
Syllabus
Video4Zephyr: From Basic APIs Towards a Full-fledged Video Subsystem - Phi Bang Nguyen
Taught by
Linux Foundation