Modeling from multiple photographs
Generally, with only one photo, we can create only incomplete model,
since several faces are hidden in the photo. If we employ multiple photos
from various angles of the same object, we can create a totally integrated
model. This example shows a model created using five photos of a
||Using these images, we can construct a 3D model which consists of 16
faces, and textures for the faces.
||Textures are retouched from the original ones to remove obstacles.
For example, the third texture from the right is modified to remove the
portion of the building standing between the camera and the target object.
It is an easy task to copy windows in the image using retouch software
such as Photoshop, because the texture has been converted into frontview.
If you have a VRML plugin with your browser, you can see the
3D demo: VRML
Last modified: May. 13. 1998