Agree with this except I believe the camera is moving from east to west between the two photos. Looking at the plane to the upper left you can just see the left hand side of the vertical fin, with the lower right hand image you can just see the right hand face of the fin. This is consistent with a the camera moving from right to left whch agrees with the apparent movement of the plane to the right of track across the ground.
I don’t think so. If the camera is moving W then the image of the plane also moves W with respect to the background. In fact the image has moved E so the camera must have a movement E (with unspecified N or S component). The view of the tail fin can vary due to the plane banking so that is not reliable.
Uh -no. Some frantic checking confirms that, with respect to the background, the object will appear to move in the opposite direction to the camera.
You can try this yourself. Move your head to the left and look over the top of your monitor. Then move your head to the right. The monitor will move to the left against the background.
Wow, if you zoom closer to the plane(s) you can see the heat distortion from the jet engines. That’s pretty cool.
You know, it is pretty obvious I am right… until I think about it carefully and I find I am still right but then… I think about it some more and it is plainly obvious you are right :smack:
I better not think about it any more. You are right and I was wrong. The observer has a Westward motion.
No problem - you should have seen me trying to prove it to myself 
And that is why when things on the ground are also moving, slowly, ice chunks in a river for example, the 3-D effect you see in a plotter will be incorrect. For example, it will show a river as a hump or depression depending on relative speeds of the ice chunks and other factors like direction of flight line.
Once you have your head around that one, you move to moving objects that are above the ground and then you speed them up and then you add the amount of overlap of the photos and the speed of the camera across the ground and then you add the relative speeds and their direction relative to the ground and then their speeds and direction relative to each other and then your head explodes…
So what do you do?
You use two aircraft and two cameras on the same flight path and altitude spaced so that the desired over lap is produced by that spacing, (GPS has made a lot of things much easier to do) and have the cameras synced and you can not only map the ground but you can accurately map things not on the ground and in motion also…
There are other tricks of the trade to accomplish stuff that makes my head explode…
This would also be consistent with the camera moving north to south, since the fin in question is pointed southeast. And the camera moving north-south makes a lot more sense when you try to imagine why a camera moving east-west would at one moment take a picture of the northern portion of the park and then a moment later take a picture of the southern portion.
Correcting what I said earlier, the camera needs to have a certain SW (actually and more precisely perpendicular to the movement of the airplane which appears to go SE) movement. It could be entirely S or entirely W or any combination of the two.
I agree moving north-south works as well but not sure what you mean by “one moment take a picture of the northern portion of the park and then a moment later take a picture of the southern portion”. How do you know there are two pictures, one of the northern part and one the southern part? (Not a snark - I’m having a real job getting my head round this.)
For people who surf the net looking at mapped images this should be clarified. The image in question is a top down image versus a bird’s eye view, which would be taken at an angle. Bird’s eye views cannot be merged in a contiguously mapped view.
Well, obviously the two spots where the plane is visible are two seperate photos, right? And we can reasonably assume that they’re two adjacent photos since they happened so close together in time. And we can probably also assume that the camera plane flies a pattern traveling due north, south, east or west. If those assumptions hold, we’ve got to be looking at one of these two scenarios: http://sdmbexample.nfshost.com/planes.png
But there’s more. We can rule out the second of the two diagrams, because the order that the pictures were taken would require that the photo plane was traveling east, but the changing visibility of the tail suggests that the photos plane must be traveling south or west, as sailor points out.
I have sometimes thought it would not be difficult to determine the height of the airplane which took some of the photos. You look at perfectly aligned city blocks and buildings and you can see the surface of the facade on the north side of the street. As you move south you can see less until at a certain point the buildings are vertically below the camera. Then, as you move further south, you can start to see the southern facades. Same thing E-W. It should be possible to calculate the height of the camera by knowing the angle between the center and the edge of the photo and the distance on the ground.
The reason the camera would have to be moving westward is that the image of the airplane has shifted eastward. So the only plausible option with our simplifying assumptions is that, as you say, the camera was moving south.
This feels like detective work!
Anyone notice the copyright mentions DigitalGlobe, among others? Some of the others do provide aerial images, as well as other products such as mapping and satellite imagery, but google only puts the copyright on there when their product is used so at least part of the image is from a DigitalGlobe satellite image.
No they’re different sizes. Any way I measure it -tail to tail, wingtip to wingtip- I get a different number of pixels. The one on the right is bigger. Here is an image of the one on the right superimposed over the one on the left. By my calculations, (assuming the correct length of the plane(s) to be 125 feet) the one from the right appears about 7 feet longer than the other.
Two planes at slightly different altitudes or the effects from stretching one image to fit with another? I really don’t know. I can’t quite imagine someone was correcting aberration and lined up everything- from the lines on the ground to the vehicles to the leaves in the tree- absolutely perfectly, and didn’t notice they duplicated an airliner.
It could still be an aerial, but the only thing that we can say for certain that goes against it being a satellite image of two planes is that airliners don’t fly formation very often.
I think that’s just a trail. If you put a straight edge on it, it curves. Why would there be a line where it was stitched together anyway? I’ve put many overhead shots together but don’t recall ever getting a line.
That scale is calibrated to the ground. Measure the length of the airplanes and it should come out to about 264 feet, when it (or they) should really be 100 to 150 ft long. Using 125 ft for the airplane length, I get about 475 ft seperation between the planes.
Ok, I just figured out that google earth can measure in feet. My last comment was measured in miles and converted, but I think it should hold. Anyway, measuring a random vehicle on the ground gives me 15 feet, so we can assume the measurement tool is close to accurate. Measuring one of the airplanes gives me 280 feet, which is way off. The planes are a lot closer to the camera than the vehicles on the ground.
Actually, bird’s eye view means the same thing as top-down. An image taken at an angle is called an oblique.
The photo plane can be moving any direction, it doesn’t have to be N,S,E,or W. There’s really several possibilities. Look at this. The plane on the left should be traveling along the green line. It has either jumped off the green line because the photo plane was moving east to west (blue), NE to SW (red), or north to south (yellow), or anything in between.
I have absolutely no doubt that it is the same airplane. It has the same marks. The likelyhood of it being two airplanes is infinitessimally small. The difference in size can easily be explained by the airplane in the photo rising in altitude between the two shots (or the much more unlikely possibility that the camera came lower in altitude).
No, it is the same airplane which has shifted with respect to the background due to parallax.
And I agree with ntucker that the camera was probably traveling south or very close to S because other directions would have resulted in the demarcation line cutting through the airplanes. A demarcation line which goes exactly E-W preserves both airplanes.
That’s all consistent, except it ignores another fact: how would the camera plane travel from east to west along the blue line, but still manage to take the picture in the upper left before the one in the lower right? This is what I was talking about in the second paragraph of the post you replied to.
But neither one has any unique identifiable marks. They both look like straight off the Airbus factory floor. It’s probably the same plane but there’s no way to say for sure.My comment about the size was only because someone said they were exactly the same size when superimposed, but I didn’t get the same results.
I should have brought this up earlier. I think everyone is assuming the images are taken exactly as simple squares that fit together edge to edge. But aerial photos always have some percentage of overlap. This would be an example of an east to west track. It’s exaggerated in that parts of the box that are outside the overlap are not to scale.
That makes no sense whatsoever. Of one thing we can be sure: the airplane was not flying backwards. Your scenario shows it is possible to take two photos like that but it makes no sense to use them in such way. None. Why would they use parts of one photo intermixed with parts of another? It makes no sense. Obviously the photos taken have certain overlap and then they are joined at the edges. But what you propose makes no sense unless someone comes aong and gives a very good explanation why it is done.