Location-Aware Images : Using Smart-Phone Capabilities To Automate Rich Metadata

The technology is here to make photography more useful and enjoyable - but the industry has yet to deliver the complete package.

The camera-phone has significant potential as an intelligent imaging device, one that adds essential data to photographs - data that can make consumer and corporate photography more useful and enjoyable.

However, after many years neither the photography or mobile communications industries have capitalized on this potential.

What are the obstacles? Why hasn't this capability come to market in a consumer-friendly package?

When might we see progress on this front?

Picture Potential

To enjoy and use a photo, you have to be able to find it.

That was hard enough when one had a shoebox with a few hundred pictures. But digital photography has brought with it an explosion in the number of photos: people who once took a few dozen shots a year now take hundreds each month. With thousands of shots on a PC hard drive, each with a generic name, no one who is not meticulously organized can find a photo they want to see. However, digital photos are computer files, and those files can contain more information than just the ones and zeroes that describe the image. This other information is called metadata. Most cameras write into each photo such metadata as the date, the camera model, and picture exposure. Unfortunately, apart from the date, very little of this information is of use to the average person - and that date is more often than not incorrect, as most users don't set the camera accurately.

JPEG image files can also have location data in the EXIF header - but because no mainstream cameras find their own GPS positions, getting this data into a picture is a troublesome, kludgey process:

With applications such as Adobe's Elements consumer image editing program or the Flickr website, users can drag and drop a tag onto a file of folders taken at one location, for example, or drag the photos onto a spot on the map. Users can later find and share the photos by searching for a location name or address, or looking at a map. Another option now coming to market is to use along with the camera a separate GPS device that notes which shots were taken where: the data and photos are synchronized and integrated on a PC later.

What's missing from that picture? Only the most numerous cameras: there are more camera-phones in peoples purses and pockets than actual standalone digital or film cameras.

These phones are actually computers: all mobile phones - not just the ""smart-phones"" with expandable operating systems - have processing capabilities better than the average desktop PC of 20 years ago.

And all that together means that the camera most people have with them - their phone - is something brand new: an intelligent camera, a smart photographic device.

It can do things with photos that ordinary cameras can not - and we don't just mean transmitting the picture via a cellular network.

Phones know who you are, where you are with GPS accuracy, and, by accessing your calendar, what you are doing and who you are seeing. The ""smart camera"" can automatically add that information to every photograph's metadata. At the very least, they can put your name and if desired your contact info into a copyright header on every shot - unlike most digital cameras. Location and event metadata would make photos much more searchable, shareable, and useable. Images can be automatically sorted, whether actually filed in appropriate folders, or 'virtually' organized based on the data and a search query. Photos can be easily accessed and searched by the picture-taker and by others as well: one family member could instantly find all the photos of gathering at a certain date and location from within another's otherwise unsorted collection of photos, for example.

On the business side, a field agent could take photos - with the smart camera stepping him through the procedure to get the necessary shots and angles, perhaps - and those photos - tagged with the shooter's name, location, assignment, etc. - would be sent, as he took them, to the other people in the company who needed to see them, and appended to the right documents, and filed in the right places.

All without intervention.

This is not an original idea; it's been around for at least five years.

LightSurf founder Philippe Kahn discussed it at a Future Image summit in 2002: the camera-phone pioneer described a device that would know for example that he was at a baseball game - based on GPS, and either the schedule in his phone or by searching the Web automatically to find events, dates, and locations - and add the game information to each photo. Since that time, GPS chipset prices have fallen drastically, and the FCC's E-911 mandate ensures that US wireless phone carriers be able to determine the location of a cell phone to within 50 meters. And yet today there is not a mainstream application that, for example, lets a Windows Mobile user click a button to have all photos embedded with the data from that day's calendar event.

Why Not Now?

The real question is, why hasn't this happened yet? Why can't the average camera-phone owner use what is now in hand, without added services or fees, to simply add information to photos taken with that camera?

In a nutshell, the problem is that the various parts of a camera-phone may be integrated in one device, but that is matter of form only, not function. There is a phone, a camera, and a computer in there, but they hardly work together easily. Yes, one can take a picture, and then use the computer to send via the phone an email with a picture attached a photo - but that is hardly all the device should be capable of. And yes, there are now applications that automatically send every photo to one's blog or webpage, but there is little else in the way of true functional integration.

What are the reasons for this lack of progress: consumer desire, industry reluctance, technical obstructions?

We spoke with executives at Adobe, ComVu, Ulocate, HP, Nokia, Sharpcast, Shozu, Verizon, and, via email, Microsoft. And we received a variety of answers.

Our report combines all the interviews in a round table style to contrast and compare the responses to such questions as:

• What would make this a practical solution?
• What are the distinctions in phone operating systems?
• What are the immediate obstacles?
• What are the intermediate steps the industry should be taking?