YoVDO

libcamera: The Future of Cameras on Linux

Offered By: Linux Foundation via YouTube

Tags

Linux Courses Image Processing Courses libcamera Courses GStreamer Courses V4L2 Courses

Course Description

Overview

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the future of camera technology on Linux in this 30-minute conference talk by Paul Elder. Delve into the intricacies of libcamera, covering topics such as CMOS sensors, color filter array interpolation, lens shading, and camera pipelines. Learn about auto algorithms, linux-surface, and UVC cameras with V4L2. Discover how libcamera handles complex camera devices, camera stacks, and per-frame controls. Gain insights into image processing algorithms, native V4L2 compatibility, and adaptation layers. Find out how to implement libcamera in your applications, including via gstreamer and Chromium. Examine kernel APIs, platform enablement, and licensing considerations for this cutting-edge camera framework.

Syllabus

Intro
CMOS Sensor
Color Filter Array Interpolation
CFA Interpolation
Lens Shading
Camera Pipeline
Auto Algorithms (a.k.a. 3A)
linux-surface
UVC Camera - V4L2
Some devices with complex cameras
Camera Stack
Camera Devices & Enumeration
Per-Frame Controls
Image Processing Algorithms (3A)
The Image Processing Algorithms
Native V4L2 Compatibility
More adaptation layers
Use libcamera
Via gstreamer
Add it to your application
Chromium (on MS Surface Go 2)
Experiment with IPAs?
Kernel APIs
Platform Enablement
Licensing


Taught by

Linux Foundation

Tags

Related Courses

Improving NVIDIA Jetson Inference Library for RPi Camera
Paul McWhorter via YouTube
Streamline Deep Learning for Video Analytics with DeepStream SDK 2.0
Nvidia via YouTube
Streaming Video Analysis in Rust Using Pravega
Strange Loop Conference via YouTube
Honey, I Shrunk GStreamer - Using Meson to Slim Down Dependencies
Linux Foundation via YouTube
How Igalia is Driving Innovation in Embedded Systems with Open Source Technologies
Linux Foundation via YouTube