Media Capture Stream Post Processing

Abstract

This document discusses use cases and requirements for post processing media streams (e.g. MediaStreamTracks, but also possibly ImageData objects and other types of Media related Typed Arrays). It is assumed this functionality will be defined in an extension specification layered on top of the existing getUserMedia API (and related APIs).

Overview

TBD - A brief discussion of the history and opportunities for Computer Vision and Video Post Processing on the Web Platform. This should include providing basic background on Post Processing and Computer Vision in general. - TBD

Use Cases

Placeholders with content TBD

Effects (mostly consumer applications)
Analysis (mostly scientific and industrial applications)

Object counting
Distribution analysis

Recognition (broad consumer, scientific and industrial applications)

OCR
Barcode scanning
Image recognition
Skeleton detection
Face detection
Face identification

Tracking (broad consumer, scientific and industrial applications)

Head/gaze tracking
Gesture detection
Gait detection
Augmented Reality
Activity Monitoring

Requirements

TBD

Examples

TBD

References

[1]: Media Capture TF Scenarios 3.3, http://w3c.github.io/mediacapture-scenarios/scenarios.html#find-the-ball-assignment-media-processing-and-recording
[2]: Media Capture TF Scenarios 8.6, http://w3c.github.io/mediacapture-scenarios/scenarios.html#post-processing
[3]: Feedback on the Media Capture TF Scenarios, https://lists.w3.org/Archives/Public/public-media-capture/2013Sep/0062.html
[4]: FoxEye - Firefox optimised/extended with OpenCV+, https://wiki.mozilla.org/Project_FoxEye
[5]: OpenVX - Portable, power-efficient Vision Processing, https://www.khronos.org/openvx/