Jump to content

Wikipedia:Reference desk/Archives/Computing/2016 September 25

From Wikipedia, the free encyclopedia
Computing desk
< September 24 << Aug | September | Oct >> September 26 >
Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.


September 25

[edit]

Does Google Earth "photoshop" trees?

[edit]

I was looking at Google Earth for coordinates 35°17′44″N 80°43′37″W / 35.29543°N 80.72707°W / 35.29543; -80.72707. When I zoomed in all the way, the tops of some of the trees didn't look natural, as if they were digitally altered. Does Google Earth do this, and why? ―Mandruss  07:20, 25 September 2016 (UTC)[reply]

Update: Ok this appears to be part of their "3D" feature, but I don't see why they would have to do it in the 2D mode. ―Mandruss  07:48, 25 September 2016 (UTC)[reply]

I've seen them do that on lawns, too. It might be a form of compression. That is, once the software recognizes a tree or lawn, rather than store every leaf or blade of grass, it just notes that region as tree or grass, then redraws it, using a standard texture, when viewed. StuRat (talk) 14:57, 25 September 2016 (UTC)[reply]
Photoshop is one particular brand of commercial image processing software; but image processing is a technique that is certainly used in many other applications, including GIS (digital mapping). In fact, GIS was one of the first applications of digital image processing. The first digital cameras were intended for surveillance satellites; the first algorithmic processing of digital imagery was intended for automating aerial images and maps; long before consumer photographers were "photo-shopping," scientists and spies were using computerized image algorithms to process aerial photographs. It's an absolute certainty that modern digital mapping companies like Google (and their data providers) use sophisticated algorithms to analyze and retouch their map imagery.
In Google Earth Engine documentation, Google explains many algorithms they make available to the public: for example, Landsat-specific postprocessing algorithms are explained. They also use sophisticated methods of machine learning to classify and segment the photographs; to discard poor imagery; to retouch, re-expose, and color-balance the pictures; and to compress and rescale the pictures. Surely Google has even more technology that could be applied to imagery that you see on their final products, even if the explanation and method is not made available to the public.
You specifically mentioned tree-tops; trees are a unique thing in aerial digital images. Find any good book on aerial photography, and you'll get chapters on methods for "foliage." There are algorithmic- and image-capture- techniques to see through foliage; to make foliage appear incredibly bright with respect to the surroundings; to use the foliage to analyze climate, water, industrial activity; to monitor seasonal changes in the treetops... the list goes on and on. Once again, it is a near certainty that "tree tops" are algorithmically altered.
Here's one of my favorite papers on digital photography: Multisensor Fusion with Hyperspectral Imaging Data: Detection and Classification. It's a sort of a review paper on how this all works. For those of you stuck on the ground, just remember - even though you can't see the surveillance planes, the surveillance planes can absolutely see you, and their powerful computers can automatically tell what type of tree you're hiding inside of, too.
Nimur (talk) 21:49, 25 September 2016 (UTC)[reply]
I put "photoshop" in quotes because I didn't mean it literally. To my amateur mind that is an approximate synonym for digital image alteration. Look at the white water tower to the east. That is not the actual photographic image of that tower, but a simulated one to enable viewing it from any angle in the "3D" mode (except from underneath, I assume). Google Earth uses this technique with "hard" structures all over the world (which is cool). You look at Notre Dame Cathedral, and what you're actually seeing is a digital "cartoon" representation of the structure. I can see why they would do that for trees as well, but I don't see why they can't show me the real and true photographic version in 2D mode. ―Mandruss  21:51, 25 September 2016 (UTC)[reply]
Indeed, Google and other major vendors sometimes use a human artist to "retouch" a complicated scene using 3D models or manual digital image alteration. I wouldn't try to speculate about any particular instance, except to mention that Google has the financial resources to hire individual human artists, at scale, for re-touching their maps imagery; but Google also has sophisticated computer algorithms that may do re-touching automatically. In still other cases, you might be seeing a complicated algorithmic or image-capture "glitch" and projecting your own interpretation onto the image defect. All of these are possible. Nimur (talk) 22:01, 25 September 2016 (UTC)[reply]

which country is lagrget in the world

[edit]

which country is lagrget in the world — Preceding unsigned comment added by Vìvek Dahal (talkcontribs) 15:19, 25 September 2016 (UTC)[reply]

By population, China. By area, Russia. StuRat (talk) 15:45, 25 September 2016 (UTC)[reply]
You may be interested in List of sovereign states and dependencies by area from Category:Lists of countries by geography and List of countries and dependencies by population from Category:Lists of countries by population as well as other lists from Category:Lists of countries. -- ToE 21:11, 25 September 2016 (UTC)[reply]
Mexico is the largest citizen (obesity) in the world. 175.45.116.99 (talk) 01:46, 26 September 2016 (UTC)[reply]
Mexico isn't even in the top 29 of obesity rates: [1]. I would have expected the US to top the list, but small pacific islands and rich Middle-East states seem to have taken the lead. ("I'll take a pass on the poi, please.") StuRat (talk) 02:53, 26 September 2016 (UTC)[reply]
Per List of countries by Body Mass Index (BMI)#WHO Data on Prevalence of Obesity (BMI ≥ 30kg/m2) (2014), Mexico ranks 32 with the US at 19 and the top seven being Pacific island nations. -- ToE 11:34, 26 September 2016 (UTC)[reply]
The US is #1 in number of obese people (Roughly ~109,000,000 followed in a distant second with Brazil at ~41,000,000 and Mexico third at ~36,000,000,000.) That is a lot of mass. We also have a list of modern great powers or List of countries with Burger King franchises or a List of rump states, a term I just learned right now. Hopefully all of these help! uhhlive (talk) 22:12, 26 September 2016 (UTC)[reply]
Why have you asked this question at the computing desk? - Champion (talk) (contribs) (Formerly TheChampionMan1234) 06:14, 29 September 2016 (UTC)[reply]