The Upoint principle

The principle behind Upoint is quite simple and elegant. 

First of all, tracking technology is used to establish position and orientation of the remote with respect to a base-station that is integrated in the projector (alternatively, the projector may be equipped with a sensor that measures its position and orientation in the base-station coordinate system). In order to be able to tell if and where the presenter is pointing on the screen, the system is missing some vital pieces of information: the position, size and orientation of the displayed images.

To obtain this information, the projector is equipped with a distance-sensor and a level-sensor. The latter is already present in most projectors, and is used to correct for the vertical keystone effect. For Upoint, it is also used to establish the 3D parameters of the optical axis which, in the definition used here, coincides with a line through the center of the lens and the center of the projected image. 

The distance-sensor is then used to establish the distance (along the optical axis) between the projector and the (center of the) projection screen. Since the width and height of the projection beam are also known from factory settings (see Figure below), the system is now able to calculate the position, size and orientation of the projected images - provided the horizontal keystone effect is negligible (note that even if this last condition is not met, additional sensors may be incorporated that also correct for horizontal keystoning, and can hence be used to help establish the parameters of the optical axis). 

With all this in place, the computer is able to construct all the necessary parameters to measure where the user is pointing, such that a cursor can be placed at that spot. 

The hardware

To demonstrate the principle of the Upoint invention, a prototype was constructed. To this end, a second-hand Fastrak system was bought on e-bay and upgraded with a TX4 transmitter (see pictures). An old TV remote, a USB mouse and a laser-pointer were disassembled and, together with the magnetic receiver, integrated in one handheld remote. Finally, a laser-tape-measure (which actually relies on ultrasonic technology) was bought from a do-it-yourself store; its reading is manually input into the system. Although the resulting system is quite bulky and wired, the final Upoint product can of course easily be built smaller and wireless.

The factory calibration

Although not strictly necessary, it would be nice if the handheld remote were also equipped with a normal laser-pointer, so that the user may use the remote to point to features on as well as away from the projected display. Depending on the point-of-aim, the laser may be automatically turned on or off. To avoid confusion when the user changes his point of aim from outside to inside the projected image, it is necessary to align the 'real laser-beam' with the 'interactive, virtual laser-beam', which must be done at the factory. It can be shown that this is possible if one is able to change the position and/or orientation of the remote whilst keeping the laser-beam fixed. To this end, an adjustable calibration platform was constructed from Lego. The laser-beam, in turn, was kept fixed by requiring it to pass through two points in space, marked on a transparent piece of plastic and on the wall respectively.

The second 'factory calibration' pertains to relating the projector coordinates to the tracker coordinates. After fixing the coordinate system (associated with the tracking device) with respect to the projector, there is a need to ascertain the orientation and position of the optical axis (by the definition used here equal to the line from the center of the lens to the center of the projected image) in the tracker coordinate system. Under normal circumstances this would be done at the factory, but since we have to make do with a separately bought projector and tracker unit, a slightly different calibration procedure will be used. For our prototype, the remote is positioned such that its laser points at the center of the projection, while its position is such that it lies on the optical axis (traced out by a second laser that is also directed at the projection-center). This procedure is shown in the figures below. This way, we can ascertain the 3D parameters of the projector's optical axis, as measured in the tracker coordinate system.

Factory calibration setup: finding the optical axis

Factory calibration: finding the optical axis

How it works...

With all the components in place a software package was written that allows the measured distance to be input manually, takes care of communication with the Fastrak module, calculates all necessary parameters and controls cursor positioning. The result is best demonstrated in a 60s QuickTime movie that can be downloaded or viewed on the Presentations page (for a free QuickTime viewer visit www.quicktime.com).

Contact us at through the Feedback page with questions or comments about this web site.
Last modified: July 11, 2007