Thursday, September 20, 2012

Control Real Indoor with Webcam Gestures and Cadence Sensor

Real Indoor uses webcam images to allow user gestures to control the Real Indoor user interface and retrieve additional information like the spinning cadence (RPM). While you sit on your spinning bike it is difficult to control the computer e.g. with a mouse and adjust power, cadence, ... - to overcome this we developed the unique gesture control.

This is build leveraging the new HTML5 features of browsers like Chrome which enable to access the video taken by the webcam of your laptop or a connected USB camera. Of course the browser will ask you if you allow access to your webcam so please confirm the question appearing just below the upper browser bar.

Below are some screenshots of the movements detected in the webcam image which then allow identifying the sensor activation and triggering the respective actions. And please remember ... all is done in your browsers javascript engine ;-)

Cadence/RPM detection:

The arrow shows the sensor where RPM calculation is triggered from. When a object (head, shoulder, crank, ...) repeatedly moves into this area the cadence is calculated and influences the power applied and so the driving speed with which the driver is moving forward on the track.

Of course you can drag the RPM sensor to the most suitable position depending if you want to measure your head movements or the rotation of the crank.

Power control:

You can control the slider on the right with the finger or hand the to adjust the power applied.

Heading control:

While driving you might want to look around. When you control the heading slider the Real Indoor driver stops and you can enjoy the beautiful scenery around you.

All controls are optional so you can choose which suits you best while spinning at full power.

User interface:

Technical background: The webcam images taken via the HTML5 "getUserMedia" javascript API are constantly monitored against changes between the present and the previous image. If movement is detected in the different sensor areas the respective control is triggered.

No comments: