As promised, the Real-time HTTP Streaming Server plug-in for UV4L is now ready! It offers a per-device Web Control Panel (screenshot) for setting the device parameters while streaming (with any V4L application) and a page for watching the MJPEG video streams (screenshot). Raw H264 streaming over HTTP is also possible. Basic authentication is also supported. For a list of options see the manual page.
For now the plug-in is only available for the RaspberryPi. It's really simple to install it, just follow the instructions provided here. For one example scroll down to Example #11 in this list.
The packages for x86 will be available in the future.
Future extensions of UV4L will be devoted to Real-time Live Streaming. To be more specific: MJPEG/HTTP Web Server and HTML5/H264 via Websockets.
As promised, the Real-Time Object Detection and Tracking for the Rpi Camera Driver is ready! This feature can be turned on and off on-the-fly while streaming.
For now the command line options allow you to pass two cascade classifiers, one used for detection and the other one used for tracking. Default settings are reasonably good for face detection/tracking at 320x240 ~15fps.
The V4L2 Camera Driver for the RaspberryPi will be supporting Real-time Object Detecton. Stay tuned!
In the meanwhile, the today's release introduces a new option known as Dynamic Range Compression (DRC) (see the manual page for more information). Don't forget to update the firmware before updating the driver.
The long-awaited Text Over Video feature (yes, Video, NOT Preview) for the RaspiberryPi V4L2 Camera Driver is ready. Update or install the driver now and try it!
From now on it will be possible to dinamically draw arbirary text lines onto the video frames at any position with any color, scale and font.
This is a simple demo showing one possible effect!
See the example #9 to know how this works.