- From com­mer­i­cal satel­lite pic­tures, the Army is putting together 3D maps of Mosul and Fallujah.
– Out of a 737 pas­sen­ger jet, Boeing is mak­ing an anti-​​sub spy­plane.
– In online chat rooms, the CIA and the National Science Foundation are hop­ing to catch ter­ror­ists schem­ing.
THERE’S MORE: With a pro­gram called FalconView, the Air Force has been cook­ing up satellite-​​generated 3-​​D maps since 1995 in Bosnia, notes Defense Tech pal CA.

“One of the best things about FV is not just that you can place items on imagery, but that you can ‘drape’ that imagery over 3-​​D Digital Terrain Elevation Data. This is what enables the hi-​​fidelity fly-​​throughs [of the area]. More cum­ber­some tech­nol­ogy, using [gov­ern­ment] imagery, has been around for years. But time and progress now allows it to get to the field…
Hi-​​fidelity imagery is not only a boon to the troops, but also to the intel­li­gence com­mu­nity. Now lower-​​priority, lower-​​fidelity require­ments (like low-​​resolution maps, envi­ron­men­tal stud­ies) can be [han­dled by] com­mer­cial [satel­lites], and not take up valu­able “national” resources as they have always done. The main draw­back to com­mer­cial is lack of assets (though improv­ing), and lack of time­li­ness. But the capa­bil­ity is great if you want to, say, get mid-​​res city graph­ics of major Iranian cities, or pro­duce mid-​​res famil­iar­iza­tion prod­ucts of Iranian nuclear/​missile facilities. 

No, this doesn’t have a damn thing to do with killer drones, pain rays, or mul­lahs with nukes. But I’m post­ing this Wired News story of mine any­way, dammit. It’s about soft­ware tools that may be able to spot the dif­fer­ence between a real paint­ing, and a slick forgery.

Scholars have had their sus­pi­cions that the paint­ing of Madonna and child cred­ited to the Italian Renaissance mas­ter Pietro Perugino wasn’t really done by him alone. But they could never be sure.
Now, a new set of soft­ware tools, devel­oped by a Dartmouth College team, seems to con­firm the art his­to­ri­ans’ doubts, show­ing evi­dence of at least four dif­fer­ent painters work­ing on the can­vas. The pro­grams’ mak­ers hope this will be the first in a long line of art authen­ti­ca­tion mys­ter­ies they can help put to rest, with code that can sort out real from fake.
“There are prop­er­ties in an artist’s pen and brush strokes that aren’t vis­i­ble to the human eye, but that are there nonethe­less. And we can find them, through math­e­mat­i­cal, sta­tis­ti­cal analy­sis,” said Dartmouth com­puter sci­ence pro­fes­sor Hany Farid, who devel­oped the algo­rithms, along with math pro­fes­sor Daniel Rockmore and grad­u­ate stu­dent Siwei Lyu.
But museum cura­tors and sta­tis­ti­cians cau­tion that the Dartmouth group’s tech­niques have only begun to be tested. Using algo­rithms to back up schol­ars’ sus­pi­cions is one thing; uncov­er­ing a fraud with just a com­puter, that’s com­pletely dif­fer­ent. And in the art world, no sci­en­tific method is con­sid­ered as sure as the eye of a sea­soned con­nois­seur.
“This is very unusual,” said Nadine Orenstein, the cura­tor of the draw­ings and prints depart­ment at the Metropolitan Museum of Art in New York. “We’re all a bit skeptical.”

THERE’S MORE: Earlier this year, Farid made noise when he unveiled his soft­ware for find­ing faked dig­i­tal images. My New York Times story on that work is here.