PF Photography Thread

NileQueen

Very picturesque.

NileQueen

Amal, I like yours too.

AlephZero

A question about image gathering from satellites in LEO, as used on Google Maps etc.

This is prompted by PF thread about a research paper where data about animals (cows and deer) was derived from images on Google Maps, and the paper claimed surprising (and not obviously believable) results.

Just for fun I looked at some Google Maps images of an area that I know well on the ground, which should have contained shown plenty of fields with cows and sheep - except there were virtually no animals visible on the images, and the few that were there (about 1% of the number I expected to see) were very blurred. Also there were also no vehicles on any of the roads, which is very improbable. (Cars that were obviously parked, near buildings etc, were clearly visible)

There was no problem about the image resolution at the highest zoom level (e.g. white painted road markings and even overhead power cables were clearly resolved), so I wonder if there is something about the image gathering technology which can't resolve moving objects, or filters them out. Long exposure times, because of the extreme telephoto lenses, for example? I couldn't find anything relevant the Web.

I'm specifically asking about satellite images here. Apparently in urbanized areas aerial photography is often used to give better resolution, and that does show road traffic, etc.

Borek

I doubt there is plenty of time to take these pictures - after all, these satellites are rather low and fast. Usually I see plenty of cars - perhaps pictures were taken early, like 5 a.m., before the traffic started?

But then perhaps pictures are taken in some counterintuitive way (three line sensors using the same lenses?) and then postprocessed to get rid of artifacts of the moving objects? Just guessing here.

AlephZero

Some images were clearly in sunlight with the sun high in the sky. From the farming activity visible they were probably taken in June. So I don't think the time of day explains the lack of road traffic.

Andy Resnick

Interesting... I never thought about this, but looking at NYC, around Times Square, using Google maps and Terraserver shows very few cars and people- much fewer than I would expect. The problem is not acquisition times- there are cars driving and people walking. The images may have been taken at odd times of day, but the shadows all seem to indicate 'normal' times.

Since the maps are generated based on many images, perhaps the images are chosen specifically to minimize the number of people/animals- images acquired in winter, for example?

Andre

Looked around a bit in my area, no traffic jams but certainly a lot of traffic at certain places.

Keep in mind that the orbits of these (this) satellite is rather rigid. So it can only make a picture of a certain area at a certain time and the next picture weeks? months? later. Then it's visible light, so there must be no clouds. For some area's that's very rare, but the least amount of clouds are in the early morning shortly after sun rise, before convection kicks in to produce cumulus clouds.

So I can imagine that having a decent picture of certain areas is a tough job in the first place, and the amount of traffic would hardly be a selection factor. Maybe that summer pictures made at something like 5-6 am, minimizing clouds, are more often successful. Probably not a lot of activity then.

AlephZero

The lack of traffic is a bit of a side issue, but it struck me as being rather odd. The "invisible sheep" would have been in the fields at any time of day or night. Granted some of the cows might be out of sight being milked in early morning and evening, though.

Borek

Wow: a year long exposure using pinhole camera.

fuzzyfelt

Wow!

Andy Resnick

very cool!

amal

Amazing. The photo's too good.

Andre

Relatively cold here with a strong easterly wind do curious things to reed sticks at the lake before the water freezes over:

This is straight from the camera:



after some tone mapping and cropping:



Edit: Same spot looking in another direction:

Andy Resnick

Maybe it's not photography exactly, but my lab class is making holograms this month, using a kit (http://litiholo.com/). I had no idea how well the kit would work, but the students made some really nice holograms and enjoyed the process, so I thought I'd give it a try myself. I decided to try making a hologram of a computer chip:



Here's the hologram of the 'bare' chip:



It's *really* bright- I had to use a polarizer to cut the intensity down enough to take a photo. The next step was to make a hologram of the chip as viewed through a microscope objective:



And here's the result:



Not bad for a toy kit!

Andy Resnick

Here's my final hologram:



The object is a computer chip, magnified with a 63mm Luminar. The holographic plate was placed near the back pupil plane, producing a virtual image at infinity. This image was taken with a 85/1.4, and since the virtual image is far away, the optics table etc. is out of focus.

There's quite a bit of optics involved with doing something like this- very non-trivial.

Andy Resnick

Today was the second half of an experiment- a professor from the Art department gave a special demo to my class. He turned a colleagues room into a camera (a camera obscura). It's really cool. Here's a couple images showing downtown Cleve-o, looking north to Lake Erie:

fuzzyfelt

Very nice!