The
AP reports that "analysis of radar data and satellite images from Sept. 3, the day Fossett disappeared in a small plane, led Air Force technicians to believe they had spotted clues to his route." That's prompted two planes and teams on foot to explore a section of the desert in Nevada where Fossett disappeared on Labor Day.
---9/24/07
On Weekend Edition (National Public Radio) this Saturday Scott Simon observed that it's telling that we can't find Fossett, who presumably wants to be found in the rugged terrain, and yet there are regularly calls of incompetence associated with the search for Bin Laden.
--- 9/18/07 ----
The Nevada Civil Air Patrol suspended its portion of an aerial search pending new leads in the hunt for aviator Steve Fossett yesterday. Maj. Cynthia Ryan reported that the patrol has searched 98 percent of the 20,000 square mile region where Fossett was last seen. Aircraft from CAP will be on stand-by at Minden-Tahoe Airport in support of ground searchers. And, the Nevada National Guard will still be searching.
-
Nevada Appeal
Also worth noting is this
blog post from from Ken Barbalace about his experience searching the images and what he's learned from the search. His list of "what to fix" for next time is required reading for anyone thinking about the future of this sort of search support.
---- 9/16/07 ---
DigitalGlobe imagery is now available via Mechanical Turk for searching. A DigitalGlobe spokesperson reports that some collects have been sent to Google. Again, it's unclear Google role in the process, but perhaps it's "gatekeeper" for imagery going to Mechanical Turk.
Also, more on the hyperspectral sensor in use for the
Mercury News:
One of the new technologies being deployed is ARCHER, an acronym for Airborne Real-time Cueing Hyperspectral Enhanced Reconnaissance.
Developed for aerial spotting by geologists and refined for search and rescue, it can "see" objects that don't belong in desert vegetation—like a piece of airplane, [Civil Air Patrol Maj. Cynthia] Ryan said. While the human eye typically detects three light bands, ARCHER can analyze 50.
Ryan also notes it's unlikely those looking at imagery on the Web will help; they don't really know what they are looking for. The plane she says is likely debris.
----9/10/07-----
Still no word from Google about GeoEye involvement.
But:
(1)
NPR notes GeoEye and DigitalGlobe provided imagery. Whether both are in Mech Turk is not clear to me.
(2)
Wired cites ONLY DigitalGlobe, though no one I've spoken to has seen a DG copyright in an image in Mech. Turk. The article also notes that 3 "versions" of the imagery exist, ideally from different passes and one is believed to have completely reviewed. 50,000 people have participated in the search via the online imagery. Each square is reviewed by 10 people - those with mulitple "votes" to look further, get higher priority.
(3)
Merchandise "commemorating" the search does not include DigitalGlobe.
----Update 9/10/07 12:20 pm EDT:---
We contacted GeoEye for information on the imagery being used in Mechanical Turk. The reply via their PR group: Their official statement is no statement and that we are to contact Google for remarks on the Fossett search. We'll try that.
--- 9/9/07---
Update (8:30 pm EDT):
AP
reports imagery is from DigitalGlobe per
AVWeb. I stick with GeoEye - per the Amazon page.
Update: 6:30 am EDT
Amazon, Google and GeoEye (I was mistaken in saying DigitalGlobe earlier; see
reference here, though I'm not aware that GeoEye supplies images to Google regularly) have
posted imagery on Amazon's Mechanical Turk so that Web users can help in the search akin to what occurred in the search for Jim Gray. Those interested can view images and note if they see something that might be Fossett's plane. Also, they can look at more detailed images in Google Earth. I don't believe that option was available last time.
C|net blogger Peter Glaskowsky offers
his experiences with the app and the search.
via
Slashdot
---9/7/07----
The AP is reporting that
indeed Google contacted DigitalGlobe for imagery. DG spokesperson Chuck Herring confirmed the company had no new recent images of the area, but would on Sunday. "We are partners with Google, so we always try to help out in any way we can," Herring said.
--- original post 9/5/07----
When I heard this morning that Steve Fossett was missing after taking off in a small plane in Nevada late Monday, my first thought was of Jim Gray. Here's another well-known person who seems to have disappeared. I recalled how the tech community jumped on its tools to try to locate Gray and wondered if that would happen again.
Apparently, it's begun. Richard Branson (Virgin Atlantic) has contacted Google about accessing imagery that might give some clues. I'm sure Google will pass on the information that the imagery folks at DigitalGlobe, GeoEye and aerial companies may have more up-to-date imagery
The Civil Air Patrol has been looking for Fossett for two days with planes and helicopters.
-
Reuters
I hope Digital Globe and GeoEye don't get harrassed into re-tasking their satellites to look for one person. If they do, I hope they do it for every lost camping and hiker. Only fair, right?? Otherwise, Houston we might have a problem.
KoS
You may have just created an entirely new and exciting market model for remote sensing, that would be value-added to the public.
Now that's using your noodle!
Anyway. Daniel....Something I've been thinking about for a few years. Whether its a specific "bird" for the task or a UAV(s). Maybe Branson and others of his ilk can fund a non-profit to do search-n-rescue assistance.
Really, I've always been bothered by the inequity or the help at all cost mentality when it's applied to a certain few. Trust me, or not, it doesn't win me many friends or invites to the social circuit. I butted heads with some in the family this summer when a family member received help from "connections". Others in the same situation(summer wildfires), didn't have the connections or even given the additional help. How is that right?
I even apply it to my personal life. I do things on my own, I don't ask for help or use any "connections". Granted I would be way, way better off if I did, but I can't live with myself if I did/do.
Terry...You have heard the ole saying, "Houston I think we have a problem"?
KoS
Continue to inspire. That is what life is about, regardless of 'connections'.
I would also run a couple algorithms over the data to see if that helps.
From loading from the direct KML link from that web site, it's extremely difficult to do this properly.
Comcast here in the Chicago area is borked. My Brighthouse connection, when I was living in Florida was about 5 times faster.
Maybe they just forgot to put some of that 'highspeed silly putty' from their many endless commercials on my connection?
I wanted to note some of the obvious things one can assume in relation to Google's involvement.
If you do a right-click properties on the Mechanical Turk images (or chips), you'll notice that they're 256 x 256 pixels. This would indicate to me that Google is taking the data, whether mosaicked from the data provider or done at Google -- and then output to KML/KMZ for rapid deployment to both the Google Earth community and to Mechanical Turk.
I do agree that most participants won't know what they're looking at, if they're not familiarized with resolution and some basic analysis principles. Also, multi-spectral often has certain common anomolies, such as color-shift -- and some pixels of interest can be overlooked by an untrained eye if they aren't familiar with that shift.
But... Anything that helps, helps. And anyone that helps, helps -- as is evidenced by some 50,000 people willing.
All eyes on deck, so to speak. (Including the professional analyists who are involved.)
Thanks for the updates on the "image chopping." I have no reason to believe regular people are looking at the multispec imagery. I believe that's coming from plane-based sensors. So far as I know only the DigitalGlobe and GeoEye imagery is in Mech. Turk for anyone to examine.
Adena
Even with a fused 3,2,1/P product, there will always be inherent anomolies present.
If what you're refering to are multi-spectral combinations with the other spectrum frequencies, then I understand the difference you're making.