WebRTC is beginning to make a significant impact on video conferencing but how do we go beyond simple calling? Why not enhance standard WebRTC video with some advanced capabilities for manipulating the media content? I recently discovered Kurento,Via this article on WebRTCHacks(http://webrtchacks.com/kurento/), and wanted to share what they are building.
Kurento is a multimedia development framework that aims to apply advanced image processing algorithms to WebRTC. It combines WebRTC with tools like facial recognition, object detection, gesture tracking, special effects, computer vision, and more.
There is awesome potential here to differentiate media applications in several verticals using these sophisticated capabilities. Here are just a few of the use cases that are possible:
– In healthcare, Kurento has been used to blur faces in videoconferences where participants want to have anonymous meetings with medical professionals.
– For security purposes, the technology can be used for reporting incidents of specific faces, violence, crowds, etc.
– In games the user’s movements in front of a webcam can be tracked to create an interactive game with no mouse or keyboard.
Kurento offers media elements and media pipelines. Media elements are the building blocks of Kurento API’s. They interact through an interface based on “pads” – exchange points of media streams among elements. In the interface, the underlying implementation is hidden so that application developers don’t have to worry about the technology behind it.
The graph of interconnected media elements is known as a media pipeline. This is typically a chain that represents a sequence of processing steps.
The European Commission has given substantial funding to develop Kurento’s framework into a PaaS (platform-as-a-service) in the cloud, called Nubomedia. This organization allows developers to create media pipelines, like encryption, transcoding, augmented reality, and video content analysis, using the Nubomedia PaaS API’s.
WebRTC is here to stay, so now the question is: how can we build upon it to create more interesting and useful applications? Kurento has taken this question in one direction, allowing programmers to manipulate and analyze the media content.