Satellite Image -KEYHOLE @ GOOGLE
 

Google has a new satellite photo program (Free Trial during Beta) - very awesome if you are into this type of thing. The program is called Keyhole and links from homepage.

Been wanting to try this... love to fool with MS Terra...

I'm installing it now. :)

Pete

It is only free for 7 days. :( , at least it was when I tried it earlier this year. Sometimes a bit slow too, but still awsome. I wonder how much better is what the spooks have. :D

Alex

Cool name choice. Knowing Google's playfulness with knowledge, it's no coincidence. Look up "keyhole" and links with remote sensing or DMA. Some of it is getting declassified. Or was during Clinton. Dunno if it still is.

Thanks for this suggestion, Bot. I found 2 interesting sites

http://www.ntnu.no/sts/content/Papers/Spyglass.html

and some unrelated info on digital imagery resolution. Check out Graham Flint, the project leader

http://www.gigapxl.org/team.htm

Tracy, I haven't forgotten about the project. I've just been overwhelmed at work. I'll probably get back to you before Thanksgiving. I should've PM's earlier, sorry.

B

Actually it wasn't named by Google. Keyhole is the name of the original program, I downloaded it sometime the spring form the developer's website.

Alex

I really wasn’t hinting about that. I saw the imagery site earlier today. I was amazed at the detail. Combined with your comment it showed that digital hardware is pretty powerful. I guess the question on everyone’s mind is to what extent remote sensing stuff can pickup a signature from enriched products, heat, gaseous emissions and so forth.... and perhaps ultimately if WMDs can be reliably discerned by remote sensing.....

Yeah, I knew it wasn't a hint but seeing your post triggered my mind.

here's a good general overview and source of useful key words and concepts for more detailed research.
http://science.howstuffworks.com/question529.htm
http://msnbc.msn.com/id/3077885/

And her's an office who's name and emblem were once secret.
http://www.nro.gov/

And here are some interesting links.
http://www.fas.org/irp/nro/

And for the truly nerdy, try ASPRS and ISPRS.
http://www.asprs.org/
http://www.isprs.org/

Then there's my favorite source of fascinating geopolitical, agricultural, and industrial tidbits.
http://www.cia.gov/cia/publications/factbook/

Keyhole was the orbiting camera do-hickey that they snatched falling film cannisters with a C-130 right?

I just checked it out, GlobeExplorer is MUCH better but is also MUCH more expensive 29.99 a month.

I will post a side by side comparison later.

Keyhole was the original code name for a satellite spy program. Before electro-optical sensors, satellites had to physically drop film, which is the origin of that plane grabbing the descending satellite. The NBC link I provided above has some great descriptions.

There are still some remote sensing applications in which film is superior to scanners or CCD arrays but of course, I don't know whether spy agencies have those particular needs.

Recently I have been able to use some aerial digital camera data and compare it to film. The two tremendous advantages of digital are that it is pretty darned flat (resulting in low spherical and chromatic abberation) and it is already in a digital format for computer analysis (no errors introduced from scanning film). Also, frame-to-frame and flight-to-flight characteristics are far and away more consistent--no emulsion/temp/processing problems.

However, the spectral richness of film is just not duplicated yet from 3 channel, discrete digital imagery. Nor is the subsequent stereoimagery as richly varied in color and texture at high resolution. But for 98% of applications, digital is going to be good enough. I just happen to work in the fringe 2%. Even so, we're going digital, too. We'll just take our time getting there in hopes that digital will catch-up when we need the bridge.

Thanks for the links, B! I’ll check them out when I’m back in the land of broadband.

While I haven’t looked, can you recommend any sites who talk more about various forms of photo imagery such as visual, UV, thermal, etc.

Also if it isn’t stepping on any confidentiality issues, what is the number of mega-pixels that is considered highest resolution for the USGS purposes?

Edit: also how big is a big CCD considered to be?

We purchase imagery based on spatial and spectral resolution requirements. So if we want say, 1 ft resolution and near IR through blue visible; 60% overlap, 30% sidelap; and acquired during a certain season, its in the RFP. It all depends on the job requirements.

lately I've been intrigued by Leica's ADS40. Its a scanner rather than a camera but they sell data in frame format. It has very good resolution and geometric properties, is reasonably priced, and very quick mission-to-analyst timing. The entire state of Florida was recently flown with the ADS40. I have looked at some areas in the panhandle in some detail and I was impressed.

For my purposes, digital frame cameras have a major drawback: bus speed. The frame is captured and must be stored quickly in order to capture sufficient overlap to get good stereo coverage. The bus speed prevents high resolution and stereo. But I have no doubt there'll be a work-around. There's just too much money to be made. I also have a problem with the digital spectrum. Yeah, its discrete and reliable, but it just isn't as rich to the human eye. This definitely affects the kind of mapping I do, which is why we still use mostly emulsion film for mapping.

But when I'm in our Cessna 185 amphib taking pictures out of the window, I love my old Fujifilm Finepix digital camera. Not made anymore, but its got about 5Mpixel, 128M card. I use mostly on the auto mode and it works great. I've abandoned SLR's, though I still have a great old Pentax (1970's) and a newer Minolta (80's) and a newer Canon (early 90's).

BTW, I see that Canon now has a digital that will accept their SLR lenses. That would be a great reason to look at the newer digitals made by old standby SLR makers.

Bot

I've spent a couple of hours looking through the links you posted. There are a lot of tantalizing clues about resolution, stated in one article as being 5” to 6” for visible light. Another article says radar can resolve 5’ or less and can actually resolve under ground to an unspecified extent. Yet another article talks about resolving plant types and detecting fire by way of spectral analysis ranging from visible to infra-red. I didn’t see a minimum resolution but the article spoke of resolving certain light types down to 400 nm. They claim 62-98% accuracy depending on the color of the item being examined. Wheat stubble was stated as most accurate.

Another whole element to this is the development of 3d imagery and of interpolating images of different views.

And yet another article said that there is a lot of remote imaging equipment that is *at least* as powerful as the Hubble, but aimed at the planet.

Teasing.


If the project I told you about is successful, I'm gonna get something in the medium format genre. I saw one reference to 16 MP back for a 6x7 cm format camera, but haven't found out the size of the CCD. My current steed is a cybershot 3.3 MP. I've done some tests of up to 12x18 prints and the quality is extraordinary overall. There is a tendency to drop off as the color approaches white or black, but the image is there, and the drop off is in the printing process. BTW, if anyone is interested, Costco will provide a 12x18” print for a mere $3 per copy. They do excellent work.

If bus speed is a limiting issue, I'm surprised that there aren’t dual imaging devices to accommodate for this failure.

Thanks again! :cool: