"Ogle Maps" Privacy Concerns are Real

Google Street View, a feature recently introduced into Google Maps, captures images of buildings, cars - and people - at street level using a fleet of vehicles, including black Volkswagen Beetles outfitted with 11-lens rooftop cameras that capture images using patent-pending technology supplied by Immersive Media. Here's a picture of one of the cars:


The results are astounding for the cities that Google Maps has already rendered images for. Here's the results of a search for a steak house near where I used to live on 9th Ave in Manhattan (note full-sized bull above the entrance):


Contrary to some reports, the resolution of many of the images rendered by this system can be quite high. Street signs, signs displaying parking rules, and license plates were clearly readable using the Google Street View zoom-in feature at many locations I visited - especially those on the West Coast.

Here's a zoom-in from an image taken at 75 E. Santa Clara St in San Jose. You can not only see what is printed on the sign ("12 MINUTE PARKING") - you can see the rivets and the dents.


This has raised issues on some blogs that marketers may use this unprecedented power to populate their databases with demographic data and further differentiate the "haves" from the "have nots" - by peering into their front and back yards.

Others have complained about privacy - about the possibilities of being spotted hiding keys in flower pots, showering, strolling in the park, or having pictures of their kids snapped while on the way to school - perhaps in front of a clearly visible address or sign.

After viewing some of the images, I agree that some modification is needed.

Example 1: Here's a Street View image of a guy picking his nose in downtown San Jose. (I put the black box over his face because I don't want to make his life any worse that it probably already is, but on Google Maps he is clearly recognizable.)


Example 2: Here's a Street View image of a woman that is currently traveling through the blogosphere (black box again added by me).


Example 3: The image below is interesting because there are so many analogs: being spotted going into an AA meeting or a shelter, entering a cancer treatment center, going to a meeting of activists, entering an outlawed church... Google removed women's shelters prior to launch of Street View because of concerns raised by women's groups.


Example 4: Here's a Street View image captured of a policeman apparently giving a driver a citation on the corner of Fulton and Laguna Streets in San Francisco (the license plate is *almost* readable but not quite).


Example 5: In possibly the most thought-provoking example in this posting (for those of us that have kids), here's a shot of some kids lying on the lawn outside their house. You can't see their faces, but on Google Maps, their address is, of course, displayed.


This image worries me more than the other examples. I cropped the photo so the address of this residential house would not be visible, but on Google Maps, the address of this house is clearly displayed - that being the point.

Obviously, Google didn't set out to create a privacy problem. In my experience, "good intentions" usually drive most engineering initiatives, and I've no doubt Google wanted to create something useful with Street View.

That said, Google has to provide a solution for this - it is only a matter of time before someone gets captured on camera making love to their legal spouse, the physical proximity of two people becomes misinterpreted resulting in the end of a marriage, or, worse, kids get targeted by criminals based on information provided by a Street View image.

The Orwellian dilemma posed by the need to simultaneously protect political and religious freedoms while protecting the community using surveillance is beyond the scope of this post.

The Electronic Frontier Foundation has placed the responsibility for fixing the issue of privacy violation at Google's doorstep. This is appropriate. Google, as publisher, is the only company that can enable a fix.

A fix would be relatively easy to implement by Google. Face-recognition technology is a fairly evolved science these days. It should not be a challenge for Google to integrate face recognition capabilities and reverse the normal logic to enable distortion or "blurring" of facial features.

License plates may be a bit trickier, but I'm sure Google, with almost $12b in cash in the bank at the end of the most recent quarter, could easily find a way of solving this issue too.

Some believe the system might already have face-blocking technology built into the system (this is a joke, folks, but there is a real image at the end of this link).

Note: If you haven't yet tried Google Maps Street View, it is, aside from these addressable concerns, an amazing technology. Check out the Bright Food Shop using this link - this is my favorite breakfast place in Manhattan. The ease of sharing this information with you demonstrates the kind of good intentions and usefulness that Google's engineers had in mind when they invented Street View.