What other features such as shape, texture and color can be suggested to classify (recognize) the three dimensional fragments in order to reassemble/reconstruct fragmented objects using their 3D digital images?
There are few issues involved in the 3D reconstruction. Plus the ones you mentioned above, I can add the camera position, orientation, focal length, and the most important one is the 3rd dimension. To project any data to the R3 space, you should exclude the z-coordinates to build accurate 3D shape. Texture and colors are a bonus for 3D reconstruction.