twin cameras processing for aerial photogrametry

[aerospace]

Hi, I wonder if a set of twin cameras pointing vertically almost 20 cm between them will work in agisoft
I know that photoscan work great solving the camera locations, but on small distance as small cm of two cameras is possible?

Our idea is to cover greater area flying at the same altitude and two cameras  looks to be better than a single more expensive and higher resolution camera

does any one have experience with this arrangement?

Regards

[aphextwin]

good idea in general. Sure it will work. But the benefit will depend on your application. We tried this with our UAV and two Canon G11 two years ago and it has two major disadvantages:

- at an altitude of >20m (which isn?t much) and parallel cameras the two images are allmost identical

- your UAV has to carry 2 cameras and the plus weight will drain the battery even more

We never tried to increase the angle between the cameras but as I understand the process, camera angles are less important than different camera positions so it would be a better idea to fly smaller rounds (for height-field).

[RalfH]

The possibility of rigid camera pairs has been discussed a while ago in another thread (http://www.agisoft.ru/forum/index.php?topic=916.msg4358#msg4358), and I still think that it could be a good idea if Photoscan had a feature that would allow processing them as pairs with a known distance.

I don't think that 20 cm camera distance would be a good idea for airborne applications though; you'd need much larger distances (e.g. cameras on the wingtips of a larger UAV). Relinquishing camera resolution and quality also is not a good idea. A single camera with high image quality and resolution (plus good overlap) will always produce better results than two poor quality cameras.

[aerospace]

I use to fly on cessnas, so is impossible to modify an airplane to mount the cameras on wingtips, perhaps some cm and less than 10 m are small distances in relation to hundreds of meters on a flight

My point is different cameras mean different values of lens-camera calibration (even those are same model cameras with same model lenses, each one have to produce specific calibration charts or data)

for 3d models to be used on animation or renders, it looks to be Ok, but for a precise surface model is an option?

twin cameras will reduce the amount of flight time  and that means $$$, so I wonder if this procedure will work on PS

Regards

[RalfH]

It doesn't matter whether the cameras have the same or different calibration parameters as long as you let Photoscan process them correctly (i.e. not force it to use the same calibration for different cameras). Photos taken by two different cameras from nearly the same point in space will not be good for photogrammetry - there have to be photos taken from different points.

I don't see how two closely spaced cameras could reduce flight time. You still need to cover a given area. Flying two cameras instead of one may increase overlap, but this can be achieved by higher frame rate.

Or do you mean that the cameras should point slightly sideways (one to the right and the other to the left)? This would increase the width of the strip photographed along one line of flight, but if base distance between these two cameras is small you will get alignment problems and increased depth error.

[JMR]

I guess the aerospace's question is not being well understood. If I'm right, the advantage he would want to exploit comes from the wider field of view as a result of the sum of that of a pair of divergent cameras that might overlap even by just a small but constant percent.
I would not expect this setup to work well in general because  Photoscan would try to orient each photo in the pair as an independent camera, not only in terms of intrinsic parameters but also exterior ones. The closeness of the two photo-centers in a pair would lead to weak resections with bad results.
The plan could work if you managed to split the projects in chunks that had no pairs inside. If the successive strips are flown in opposite directions then you could form chunks with every two successive pass halves.
>LLLLLLLLLLLLLLLLLLLL
>RRRRRRRRRRRRRRRR.     TURN
==============}chuk one with R photos
RRRRRRRRRRRRRRRR<
LLLLLLLLLLLLLLLLLLLL<
==============}chunk two with L photos
>LLLLLLLLLLLLLLLLLLLL
>RRRRRRRRRRRRRRRR.     TURN
...etc. This planning would require ground control for every two rows... And that sounds bad.
just thinking, a better solution could be to alternate trigger of left an right cameras so they did not fire simultaneously, this would work well if you let left an right camera overlap over 30% a variant that could also be possible with one single camera capable of tilt from left to right in cycles... Mmmm sounds feasible, but still much more complicated than shortening the focal length of your camera.
By the way, what cameras are you flying with?
Regards