|
The context photography application is now working on the
Nokia 6600 and 6630 phones, thanks to Mattias Rost and Panajotis
Mihalatos. We have also recently completed a user study with people
running the
context
photography application on their own camera phones during a period of 6
weeks. We recently demonstrated the final prototype and presented the
study's
results in Oslo, Norway at the NordiCHI 2006 conference
(14-18 October 2006).
Context pictures taken by Elizabeth Goodman with our
application running on her Nokia 6600 phone, November 2004,
San Francisco, USA.
Digital photography, context awareness, participatory design, mobile
users, Lomographers, digital media, alternative photography,
environmental photography
When taking a picture, would it be possible and interesting to capture
something in addition to the visuals of the scenery? We explore if e.g.
sound, temperature and pollution could be used as parameters in a
digital camera and how these phenomena could be "visualised" or
"represented" in a picture. The fundamental idea is to go beyond the
traditional parameters (light, speed and focus) used in both analogue
and digital cameras, and widen the concept of what a camera can
capture. Digital technology opens up for new dimensions that can affect
photography! This means that digital cameras no longer have to resemble
their analogue counterparts. Unlike the editing of a picture in e.g.
Photoshop, we want the (new) parameters to affect the image in
real-time.
We have collaborated with a focus group of open-minded and explorative Lomographers, as well as with
other dedicated amateur photographers. They were not primarily our end
user group, but a source of inspiration (and also participants in
design sessions) when designing the camera prototype.
Results from a user study performed with the latest version of
the prototype are described in the following full paper published in
October 2006 at the NordiCHI
2006 conference in Oslo, Norway:
"More Than Meets the Eye: An
Exploratory User Study of Context Photography" [pdf]
The latest version of the context camera running on camera
phones is
described in the following demo abstract, published in September 2005
at the UbiComp 2005
conference in Tokyo, Japan:
"Context Photography on Camera Phones"
[pdf]
The interaction prototype and its use is presented in the
short
paper "Context Photography: Modifying the Digital Camera Into a New
Creative Tool" which appears in Extended Abstracts of CHI 2004. The short paper can be
downloaded here:
"Context Photography: Modifying the
Digital Camera
Into a New
Creative Tool" [pdf]
The initial ideas on Context Photography (Context Aware Photography)
was presented as a research sketch at Designing
for User Experiences (DUX) 2003 in San Francisco, USA. Download the
paper and the presentation here:
"Capturing the Invisible: Designing
Context-Aware
Photography" [pdf] (Large file!)
DUX presentation [ppt]
The current context camera prototype was developped for a longer-term
user study and consists of a software
application running on camera phones. It currently
runs two standard camera phone models, the Nokia 6600
and 6630. The application uses the device's own hardware
(lens and microphone) to which it connects with custom-made
code. It is programmed in C++ using the graphics
library GapiDraw, a multi-platform graphics library
available for various handheld devices. The application
measures sound level and computes the power of low,
medium, and high frequencies. Movement is retrieved as a
vector field at different points in the picture, using
optimised algorithms from the image processing
programme Optica.
The user can capture
images, see the resulting photographs, save them, and
browse through them the same way she can with a regular
camera phone. The user can also choose anytime among the
four different sets of graphical effects, calibrate the sound
and motion sensitivity, browse a gallery of taken pictures,
and delete taken pictures by simply using the phone's
middle joystick. In picture taking mode, the original image is always
visible
in the viewfinder. The resulting image with applied effects
is made visible once the picture has been taken. The user calibrates
the sensitivity of the sensors
individually, such that she can let one set of effects be
strongly affected by the sound level, whereas another set of
effects might not be affected at all. The calibration feature is
implemented differently in two
versions of the application: in the first version, the
calibration is part of the main menu and the users see the
changes in sensor sensitivity with numbers, while the
effects are only visible once the picture taken; in the second
version, one enters a special calibration mode where the
effects are continuously visible, and goes back to take
pictures in the main mode where the effects are not shown
right away.
Effects combinations have been re-designed based on input from previous
prototype evaluations and are now the following:
1. Colour shadows: Traces of coloured shadows
follow movement; the colour of the shadows
changes with the frequency spectrum of the
surrounding sounds.
2. Zoom: The part of the picture with most movement
is zoomed in, and rendered as a transparent layer
on top of the non-affected image; the amount of
transparency is determined by surrounding sound
level.
3. Pixel: Small white dots follow movement as a
decaying trace; the size of the pixels in the picture
is proportional to the surrounding sound level.
4. Waves: Movement creates waves in the image,
making it look like a dense liquid. As in 3, the size
of the pixels in the picture is proportional to the
surrounding sound level.
Visual effects in the cameraphone prototype
Longer-term user study (summer 2005)
More information about the user study coming soon!
The following describes the implementation of our previous prototype
that had been developed for testing simplified yet
realistic use, as well as exploring issues related to real context
input. This was done in collaboration with Panajotis Mihalatos who
implemented a simple software prototyping platform and a set of
effects. Vectorial movement and sound level were already chosen as
input.
The effects were grouped in combinations of one "movement" and one
"sound" effect.
The prototype was implemented on a Tablet PC, with the screen acting as
a viewfinder, and all processing was performed by a C++ software
program. A webcam served as a lens, and a small mouse taped on top of
it was used as a trigger (see picture below). A condensator microphone
connected through a small pre-amplifier, measured the sound level.
Movement was retrieved as a vector field from the differentiation of
subsequent images captured by the webcam. Input was calibrated to
default values corresponding to a normal image without effects. Effects
combinations were the following:
1. Small white dots followed the movement as a decreasing trace + pixel
size increased with sound level
2. Traces of coloured shadows followed the movement + the rest of the
colours evolved towards a grey scale with increasing sound level
3. "Swimming-pool" effect + colours evolved towards a grey scale with
increasing sound level
4. Extreme zooming on movement + colours evolved towards a grey scale
with increasing sound level
The user pointed with the webcam, saw the image and its real-time
effects on the viewfinder, took a picture by left-clicking on the
mouse, and changed effect combinations by right-clicking. When a
picture was taken, an audio feedback was heard; the image froze a
couple of seconds on the screen, and saved as a JPG file.
 
Previous camera prototype and pixel effect (click on image to see
quicktime movie showing pixel movement effect)
Evaluation workshops (December 2003)
The aim of the workshops was to evaluate our concept and get feedback
from users trying out the interaction prototype. We chose to involve
two different groups of users in two different settings, in order to
get diverse feedback. The participants of the first workshop were two
of the previous lomographers, and one amateur photographer. Those of
the second were two high school students interested in photography.
Instead of staying in a lab, we chose to conduct the workshops in
everyday settings: the local central train station and a high school.
The workshops were conducted in similar ways and both lasted about 2
hours. The sessions were documented with video- and audio- recordings,
written notes, and photographs. We started by explaining context
photography, using the concept prototype to illustrate the idea and
make it more tangible. Then we let them try out the interaction
prototype in an informal way. We neither gave specific tasks to fulfil
other than to take pictures, nor restrictions on time. Finally, we held
a semi-structured discussion about the concept of context photography,
their experience of the prototype, general characteristics of the
effects and suggestions on improvements. The photographs taken with the
prototype served as references during the discussion.
Workshop 1:
The top three pictures show the Lomographers taking pictures with the
prototype camera at the train station in Göteborg. The middle
pictures are examples of effect # 3, 2 and 1, and the bottom three
pictures are examples of effect #3, 2 and 2.
Workshop 2:
The top three pictures show the high school students taking pictures
with the prototype camera at their school. The bottom pictures are
examples of effect # 2, 1 and 4.
  |
