Is there a way to digitally capture a 3d point of view image of all directions in a sphere? Your eyes take in light from the fields of view, approximately 160 degrees but you cannot capture the other 200 degrees at the same time or can you?


Yes you can.

The technique you use depends on 2 things. If you need a high resolution option or you want a realtime one.

For the high resolution option you can take a series of shoots rotating the camera and after that stilching it on a special program.

Taking a spherical projection

You can use Hugin that its free.

You need some technique to do this. Here are some notes I did sometime ago: http://www.otake.com.mx/Apuntes/Imagen/EnviromentMaps/

If you need exact aligment in an interior shoot you need a special head Do I "need" a panoramic head to shoot 360 panoramas?

For a realtime shoot, lets say people is walking arround or you need video you can take a shoot from a ball mirror. That projection is called a Hemispherical probe.

You can transform that projection in a spherical one but the resolution, specially on the borders is very low.

You can make an array of cameras too. sometimes the images will not align very well. Like google street view.

The 3D part

You can project this into a 3D sphere to make a panorama or environmental sphere, which is 3D.

But the position of the point of view is exactly the center of this sphere.

If you need to make a 3D model of that you need multimpe shoots or a moving video so a program can calculate the diference and starting to create a 3D model.


It is possible to create spherical 3D views with a lot of cameras in precise arrangement, post processing and a VR headset. See this blog post by the esteemed Vi Hart for the gritty details: http://elevr.com/stereo-polygons/


It's possible with multiple cameras taking a shot at the same time.

One option is a ball with cameras facing outwards in all directions.



It is possible to capture a full 360 x 180 degree panorama (allowing you to see in every possible direction), using either multiple cameras or specialized reflectors. However it will not be "3-D" in the sense that the image will contain no depth information, you couldn't turn it into a 3D model of a room and rotate it etc.

  • Images actually do have depth information. After all, we humans have a certain perception of depth when we look at an image, so this information is in there somehow. We make assumptions about what parts of an image belong together to form objects and what objects are in front of other ones Here's an example of a computer doing that This is a fairly complicated algorithm, but I guess if you think about it in simple terms: the blurriness of a part of the image is via optics related to its distance from the camera = depth
    – null
    Jun 8 '15 at 14:20
  • @null you can guess at depth, but that's it, there's no information there, from a technical standpoint. The best example of this is that it's impossible to tell the difference between a photograph of an object (that has depth), and a photograph of a photograph of an object (which has little-to-no depth). The reason you can't tell the difference is because there is insufficient information in the image.
    – Matt Grum
    Jun 8 '15 at 15:44
  • take a look at the link. If there was no information, I don't think this would be possible.
    – null
    Jun 8 '15 at 18:12
  • @null You can make guesses, certainly, which is what that paper is doing, but that's not the same as having actual 3-d data. The correspondence between two images taken from different positions does contain 3D information.
    – Matt Grum
    Jun 9 '15 at 12:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.