# Conversion of 2D images to 3D

I'm progamming in VC 6.0 for some time now and want to learn and develop image processing project for my college labs and with reguard to that, i want 2 know as 2 how to convert 2D images (standard types images) to 3D format.

Vivek.

• : I'm progamming in VC 6.0 for some time now and want to learn and develop image processing project for my college labs and with reguard to that, i want 2 know as 2 how to convert 2D images (standard types images) to 3D format.
:
: Vivek.
:

If you mean convert 1 2D image to a 3D object then there isn't a general solution because there are infinitely many ways of doing it, and there are definitely more than one meaningful ways to do it giving different results. Maybe you might want to specify more specifically what you have in mind. Of course, know that it will never be "perfect". E.g. what's BEHIND the image?

Converting a series of related images into a 3D image is more specific, but to give you specific help one would need to know the format and intent of those images. Assuming for example a simple setup of a series of cross-sections a simple solution would be to use Voxels. Just grid the image if the image is black (we'll assume monochrome black==occupied), then add that Voxel at that grid coordinate using the index into the streams as one of the axis.

"We can't do nothing and think someone else will make it right."

• Fundamentally, what you are asking for is not possible, but there are always approximations to the impossible.
Another solution is image based rendering ( check www.cs.unc.edu they do some nice research ). I won't get into the details. But you can take two images, perform a specific projective transformation on them and then interpolate to get a "3D image". This effect was made famous in the Matrix, etc.
ks

: : I'm progamming in VC 6.0 for some time now and want to learn and develop image processing project for my college labs and with reguard to that, i want 2 know as 2 how to convert 2D images (standard types images) to 3D format.
: :
: : Vivek.
: :
:
: If you mean convert 1 2D image to a 3D object then there isn't a general solution because there are infinitely many ways of doing it, and there are definitely more than one meaningful ways to do it giving different results. Maybe you might want to specify more specifically what you have in mind. Of course, know that it will never be "perfect". E.g. what's BEHIND the image?
:
: Converting a series of related images into a 3D image is more specific, but to give you specific help one would need to know the format and intent of those images. Assuming for example a simple setup of a series of cross-sections a simple solution would be to use Voxels. Just grid the image if the image is black (we'll assume monochrome black==occupied), then add that Voxel at that grid coordinate using the index into the streams as one of the axis.
:
:
: "We can't do nothing and think someone else will make it right."