Role | Artist | Researcher | Software Developer |
Tasks | Research | Installation Design | Creative Coding |
Client | UCSD Interdisciplinary Computing and the Arts BA Thesis |
Installation space layout design.
Installation space as implemented at UCSD Annex Gallery in June, 2009.
Projection 1 (Side Wall Face Distortion Video Feed) - Real time distortion of the participant's face via infrared camera, face detection and motion detection. The more intense the radiation, the more distorted and chaotic the video feed. When radiation passes a threshold, the viewer's face is rendered with a skull image, denoting potentially hazardous levels of radio frequency radiation exposure. These visuals are projected on the front wall.
Projection 2 (Floor) - Generative cellular automata simulation based on Conway's Game of Life that reacts to electromagnetic radiation. Rule sets change based on the magnitude of EM fields, creating different environments and cell behavior. This is projected on the installation floor space to create an slowly changing ambient environment.
Projection 3 (Side Wall) - Visuals of hertzian space disturbances and "hotspots" projected on the
installation
side wall. Changes in electromagnetic field intensity are reflected through
the emergence of globular hotspots that quickly spread and influence the color
and shape of nearby zones, creating a constantly changing colorscape. Like
a radarscope, the visualization updates itself in a circular fashion with
new hotspots emerging around a spinning sweep indicator (a randomly color
changing "organism" comprised of a large ellipse with a trail of smaller ellipses).
The speed and direction of the sweep changes based on radiation levels to
reflect the dynamism of hertzian space.
When there is no detected radiation, the indicator moves slowly in anticlockwise fashion.
When the indicator starts to move clockwise it means that a burst of radiation
has been detected. At this moment and from the position of the indicator, there
is also an explosion of expanding transparent ellipses and a string of ellipses moving on a curved
path. If there is constant radiation, the clockwise rotating indicator will
continue to produces these explosive elements. As radiation levels increase,
the sweep indicator will rotate faster, making the explosion of expanding
ellipses appear further away from the indicator and towards the edge of the
2D colorscape. The overall effect is a more agitated and dynamic visualization
of an active space.
Projection 3 (Side Wall) - Radarscope inspired visualization and spinning sweep indicator.
Sensors configured to measure the EM fields:
Top Left: Alan Broadband Co. ZC 190 B meter + webcam
Top Right: Cornet D-15 Meter + webcam combo driving face detection projection.
Bottom right of image is side view of face detection webcam.
Bottom Left: Cornet ED-25G Meter + webcam driving "hotspot" visualization.
Bottom Right: Converted infrared webcam with custom AA powered IR led module.
Technical Details
Hertzian Explorer relies on an unconventional configuration of existing
technologies to connect with the invisible spectrum. EMF meters are
mounted to webcams to perform computer vision analysis of the meter's
displays. Motion detection or brightness/color tracking techniques
are used to convert and transfer the visual data from the meter's
display into numerical data fed into artist written programs running
on hidden computers to help generate visualizations of hertzian space.
In my openFrameworks face distortion program, if the signal strength bar
on the ED-15 meter illuminates and flashes due to the prescence of
radiation, the computer vision algorithm doing motion detection will
convert the changes in illumination to a number that becomes a visualization
parameter. With brightness/color detection, the brightness of the LED radiation
indicator on the broadband meter is converted to a value usable by the Game of
Life based Processing visualization. Lastly, to help generate the side wall
"hot spot" visuals, my Processing program tracks the y-axis position of the
brightest pixel in the webcam video feed of the ED-25's display. This position
data directly corresponds to signal strength.
Two webcams are used to implement the face distortion feature.
As described above, one webcam is used to transfer data from the meter
into software. This camera is doing motion analysis on the RF meter's
display to quantify the amount of radiation into data usable for the
visualization. A second webcam performs face detection. This specially
converted infrared camera is attached to a custom made IR led light for
better illumination and face detection and tracking performance. When
the face of a cell phone user/victim is detected, a distortion effect is
applied over the subject's face area. Using data collected by the first
webcam, the distortion modulates its intensity based on the severity
of RF radiation.
All images and content copyright Xuan Li 2025+