Workday mornings are best spent sitting cross-legged on the recliner
with a mug of hot tea and the day’s Wall Street Journal. This is
the best vantage for reviewing sonograms and other family portraits.
The current digital camera is finished. Never mind the new imaging
technology that captures color as accurately as 35mm film. This goes
way beyond that. It makes you appear a lot smarter than you actually
are.
Digital Camera II has the following: GPS, JPEG 2000 encoding, audio
playback, and an image analysis engine. Some versions of these look
like binoculars, since they take stereo pictures. Others are like
the current ‘night vision’ goggles, they can photograph in the infrared,
or in extremely low light conditions (what we would normally consider as
moonless nights).
The purpose of the GPS is to combine the picture with it’s exact location
on Earth and a precise time of capture. Other camera settings, such
as shutter speed, focal depth, and so forth are also recorded.
You take a picture of the Grand Canyon from a particularly popular outlook,
and hundreds of others do likewise at different times and dates.
All of you publish these pictures on Web sites, with the GPS and camera
settings intact. Search engines are able to locate all pictures taken
from this location in the general direction of the canyon, and can then
create a movie montage of the view from that outlook over months, years,
or decades. This is interesting when watching forests, glaciers,
or other natural processes that evolve slowly. The JPEG 2000 format
stores GPS and setting data combined with the picture.
Note: |
The JPEG committee is working on a JPEG 2000 specification
now. Their Webmaster, answering my query about the GPS coordinates,
indicates that one file format [EXIF] already supports it. The specification
may provide for a ‘general’ labeling scheme, where files can be saved with
any number of text strings, put there from the original point of capture
or later by editing tools. |
The audio playback gives you a running narrative of the things you are
photographing. You aim the camera at Notre Dame in Paris and shoot.
The ‘Paris Guided Tour’ flash memory card first looks up the general location
of your shot, finding it in proximity to a famous landmark. Assuming
you have your headphones on, it begins to play out a description of the
building, including stories, history, architectural features, purpose,
and so forth. If you’re traveling through the lower Mississippi valley,
you can trigger this output without taking a picture, just listen as you
drive through areas of historical interest.
The image analysis extender slot is what makes you into Einstein. You
photograph a tree, which is recorded into flash memory as usual.
The image analysis engine then parses the image, finding that the picture
is of a 100 year old oak tree, and certain parts of the tree are infested
with a particularly notorious pest. The summary of this information
is written out to another file in the same flash memory module. At
this point, you have at least a hint that something is wrong with your
tree, expertise you acquired without hiring someone to come look.
This can be taken a bit farther, in that the analysis software can direct
you to a closer viewpoint, where you take another picture. In the
more distant snapshot, the analysis engine was only able to identify something
that seemed out of order, as you ‘home in’ on a particular region it gets
progressively more precise about what is wrong.
This idea can be expanded in a number of dimensions: identifying specimens
of various animal and plant species; identifying rare vehicles, aircraft,
or boats; estimating the area of a surface that needs to be painted; or
even hinting to the photographer setting and angle changes that will improve
color balance or reduce glare. Particular ‘expert’ modules are downloaded
from respective Web sites for $19.99 cheap, not including the cost of re-programmable
media.
The combination of historical narrative (particularly local historical
narrative), taxonomic classifications, and instructional walk-throughs
creates a vast opportunity space for content creators. If you are
an expert on horses you are able to ‘can’ that expertise and sell it to
digital camera users the world over. Same goes for building repair
estimating systems, collectible identification systems, ‘bird watching’,
and feedback driven instruction. If a book has been published on
the topic, this is where it can be installed as a real time system, and
all that knowledge can be tested against a real world recognition system.
The camera market as it is currently constituted is saturated, so prices
have been dropping precipitously and resolutions are now, at five megapixels
plus, excessive for all but the professional. Five million pixels
is roughly an 1800 by 2700 pixel array. The human eye cannot resolve
dot features that are smaller than 400 dots per inch, so at this pixel
density the output is 4.5" x 6.75". An 8" x 10" portrait, using this
rule, requires a 3200 x 4000 pixel imaging element (12.8 megapixels) and
36 megabytes to store one snapshot. A generator is helpful to provide
enough power to the camera so that it can finish recording the image to
flash memory.
A whole bunch of people you never heard of can now ‘get in your face’.
The Texas A&M professor who is an expert in fungi makes up an Internet
download so that people in various parts of the world can identify what’s
growing on their tree in their backyard. When the application can
tell it’s a fungus but can’t tell which one it is, the user is invited
to send the picture back to A&M, and the professor can figure out whether
there’s something out there no one has seen before. If this is true,
he can name it after you, or your prodigy child or geographical region
if you’re not into becoming a permanent part of the botanical catalog.
Some of the information providers have ulterior motives. If you
have a ‘yard care’ expert in your camera and it sees brown patches in your
yard, someone will be at your door in about ten minutes to sell you something.
How exactly they were notified isn’t clear, since you didn’t hook the camera
up to your PC and there’s no wireless... oops.
|