Let me explain what I mean by lack of data:

1. If you do WiFi positioning, you need several fingerprints (i.e.
lists of access points) in order to get enough information (and
accurracy): Who's going to stumble that for whole public buildings and
cities? (I'm going to launch a crowd sourcing approach with my
IndoorWPS).

2. Picture database: I've experimented with image recognition similar
to www.kooaba.com: That needs lots of pictures and technology is far
away from consumer market.

3. Existing data like OpenStreetMap or Wikipedia (see
www.mobilizy.com/wikitude.php) don't have yet really compelling data
for AR.

-S.

P.S. Here's another app with an AR Identity idea: http://bit.ly/ZbZYO
- quite frightening to some more privacy oriented people.


2009/8/12 Tom Longson (nym) <[email protected]>:
> I don't agree with the lack of data argument. There are many ways to
> use existing data to create compelling AR.
>
> Precision though, still is such a big limiting factor. How can I build
> an app that does contextual overlays that are anything but based on
> image processing? Locative technology has yet to make the next big
> leap. It's good for navigating you to a restaurant, but useless for
> navigating you to a table. That's the big difference in my mind.
>
> Cheers,
> Tom Longson (nym)
> ------------------------------
> http://tomlongson.com
>
>
>
>
>
> On Tue, Aug 11, 2009 at 4:54 PM, Mike Liebhold<[email protected]> wrote:
>> Stefan Keller wrote:
>>>
>>> To me the problem is 1. Bad precision and 2. Lack of data.
>>>
>>
>> Some quick thoughts on lack of interest on geowanking based on the two
>> problems you mention above:
>>
>> 1."Bad Precision
>>
>> We can safely anticipate precise positioning is on the way to being solved,
>> at least for outdoor urban apps. Google, Microsoft and Nokia have all
>> announced they are work on precise positioning of handsets, by matching
>> significant points in images with a stored datablase of 3D points. A few
>> weeks ago, at the GML Geoweb conference in Vancouver, Michael Jones, from
>> Google showed a  3D model google guys mashed up from flickr images using
>> something like MS photosynth. Aside from Earthmine, which I already
>> mentioned,  I know of several unnanced projects to simial things, with much
>> greater precision.
>>
>> In any case, A lot of AR will work fine simply using a GPS and compass.
>>
>> 2. "Lack of Data"
>>
>> What's new?! In 2003 when Joshua Schachter launched  geowanking, there was
>> almost no geocoded web data, aside from his home grown GeoURL.  That didn't
>> stop anyone on this list from experimenting. About the same time Chris Goad,
>> Jo Walsh, Dan Brickley and others created a geo RDF,  ( IFTF.org)  Chris
>> implemented  a  working demo  for us, at IFTF.org of  geoRDF web annotations
>> with integrated vector shapefiles from  the National Park Service. This
>> demo, in '03 or '04 was the first full blown demo of a geospatial web.
>>
>> Since then,   People on  the Geowanking list including, Mikel Maron, Andrew
>> Turner, Alan Doyle, Raj Sing, Sean Gillies, Chris Schmidt,  Carl Reed, Ron
>> lake and others,  designed and implemented  GeoRSS which has resulted in
>> countless geocoded points, and sincel then, as we all know google has
>> promulgated KML which has resulted in milions more geocoded points.
>>
>> Summary: It's time to start again. We need a simple way of geocoding
>> augmented reality, and many ways of processing searching and  and viewing
>>  geo AR.
>>
>> Historically, the geowanking community, has been the crucible of innovation
>> for a geospatial web.  ( see above)  Maybe five years ago, everyone was out
>> on the street, and hungry enough to build a totally new ways to  experience
>> spatial data., and maybe now everyone is gainfully employed and too
>> distracted to launch a second perhaps even more interesting spatial web,
>> viewable throught the viewfinders of our phones, and later on, perhaps in
>> eyeglasses.  If that's the case, then we can all just sit back and let
>> google build the whole thing without us. Clearly their latest UI
>> implementations in Streetview are part way there.
>> see:
>> http://google-latlong.blogspot.com/2009/07/zoom-photo-navigation-in-street-view.html
>>
>>
>>
>> ~ Mike
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>> We experimented with indoor enviromnents based on our Indoor WPS
>>> approach and server (see http://gis.hsr.ch/wiki/IndoorWPS). These are
>>> some screenshots of an Android prototype:
>>> http://dev.ifs.hsr.ch/indoorguide4android/wiki/WebDocu .
>>>
>>> So according to Mikes explanatory selection my answer is:
>>>
>>>>
>>>> (X) d. Already into AR 3dgeo, not saying much about what we're doing
>>>> until it's ready.
>>>>
>>>
>>> Yours, Stefan
>>>
>>> P.S. BTW I also was expecting bit more of response about my recent
>>> thread about "Better auto-discovery in the Geo-Web through "see" and
>>> "see also" links?" since I thought geo-search technology could be of
>>> interest to geowankers. I still did'nt give up.
>>>
>>> 2009/8/4 Mike Liebhold <[email protected]>:
>>>
>>>>
>>>> wow!  I didn't expect many replies  from busy people to my questions
>>>> about
>>>> handheld viewfinder AR views of 3d geodata,  but given the staggering
>>>> potential for AR as a radical new kind of mapping is mind-boggling, I am
>>>> a
>>>> bit surprised that  in the 24 hours of my post, this normally verbose
>>>> community has nothing to say.  Aside from a mild suggestion by Sean
>>>> Gillies
>>>> of "democratizated service-oriented architectures of curation"  no one
>>>> else
>>>> offered any thoughts at all about  viewfinder AR views of 3d geodata.
>>>>
>>>> We can only guess why the silence:
>>>>
>>>> ( ) a. Mostly into map views of geodata, not  very interested in
>>>> viewfinder
>>>> AR 3d geo.
>>>>
>>>> ( ) b. AR 3dgeo is interesting, but haven't thought much about it.
>>>>
>>>> ( ) c. Have some ideas about AR 3dgeo, looking into it, waiting for apis,
>>>> to
>>>> experiment
>>>>
>>>> ( ) d. Already into AR 3dgeo, not saying much about what we're doing
>>>> until
>>>> it's ready.
>>>>
>>>> ( ) e. Have some ideas a bout  AR 3dgeo, but am too fr*&#n busy to
>>>> respond
>>>> to random listserv e-mails :-)
>>>>
>>>> whatever. AR looks like a fun, open frontier in both geoscience and
>>>> geohacking, let's hope the foss gang gets there before goog & apple own
>>>> it.
>>>>
>>>> - Mike
>>>>
>>>>
>>>> Mike Liebhold wrote:
>>>>
>>>>>
>>>>> This is a humble request for geowanker mappers and geocoding web artists
>>>>> alike to join me to quiet our 2d cartographic minds for a minute to
>>>>> engage
>>>>> in a little thought experiment about  handheld views of 3d geodata:
>>>>>
>>>>> Context:   The realtime tweetsphere  http://bit.ly/rZncR and
>>>>> youtubesphere
>>>>>  http://bit.ly/UwQ3u are alive with news of handheld AR, [agumented
>>>>> reality] capabilities & apps, and services. Besides layar and wikitude
>>>>> on
>>>>> android, there are already a number cool hacks for jailbroken iPhones,
>>>>> and
>>>>> veiled confirmations from apple that the next rev. of the iphone os dev
>>>>> release for the 3gs will support AR app integration of gps, compass, and
>>>>> graphic overlay of video cam views.  The viewfinder is becoming a new AR
>>>>> 3d
>>>>> web browser.
>>>>>
>>>>> AR is here, if not today, tomorrow, but i suspect our geosphere is not
>>>>> ready.
>>>>>
>>>>> questions for fellow geowankers:
>>>>>
>>>>> 1. What kind of geodata and locative media will be most useful or fun to
>>>>> see geopositioned in 3d in our handheld viewfinders.
>>>>>
>>>>> e.g.  floating labels of things, animated directions, emergency alerts,
>>>>>  news and ads to filter, cartoon games, visible clouds of wifi signal
>>>>> strength, visible sensor net readings of air quality.. visible entry
>>>>> into
>>>>> another colored polygon
>>>>> of cartographic meaning, etc.
>>>>>
>>>>> 2. How will we search, view, create, and serve open 3d AR geodata?
>>>>>
>>>>> 3. What is the  prospective FOSS stack for 3dAR geoservices?
>>>>>
>>>>> e.g.  starting at the top with a 3D firefox or open layers equivelant
>>>>> client down to a cloudwide RESTful deep geocoded web of linked 3D
>>>>> geodata?
>>>>>
>>>>> Just curious, what people are thinking these days . . .
>>>>>
>>>>> Mike
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Geowanking mailing list
>>>>> [email protected]
>>>>> http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
>>>>>
>>>>>
>>>>>
>>>>
>>>> _______________________________________________
>>>> Geowanking mailing list
>>>> [email protected]
>>>> http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
>>>>
>>>>
>>>
>>>
>>>
>>
>>
>> _______________________________________________
>> Geowanking mailing list
>> [email protected]
>> http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
>>
>

_______________________________________________
Geowanking mailing list
[email protected]
http://geowanking.org/mailman/listinfo/geowanking_geowanking.org

Reply via email to