Stereopairs

Two images of the same area, taken from different points of view are called a stereopair.

Stereopair of Tongerlo, Belgium, 17-05-1999. The overlapping areas are marked in the red rectangles.
Source: Ghent University

In a stereopair, there is always an overlap (see red squares), where stereovision is possible, because of the present parallax. Outside these regions, you are not able to see in 3D.

Viewing systems

To get these two images merged together in a 3D image, the observer has to see the pair of images through a viewing system called a stereoscope. The stereoscope is a tool which guarantees that each eye sees only one of the stereopairs. Each eye thus sees a different image of the same scene, each image taken from a slightly different point of view. When the brain merges the two images together, the observer experiences a 3D vision of the scene. This can be realized manually or digitally.

Manual vs. digital stereoscopes.
Source: Ghent University

How do you obtain stereopairs?

Aerial photography

When scientists want to map out a specific area, they first create a flight map. Flight lines are laid out on this flight map with a spacing that allows the photographs to cover a common strip of ground. This overlap between flight strips amounts to about 25% of the width of the area covered by the photograph.
Each photograph in the flight line covers an area which overlaps the area covered by the previous photograph by about 60%.

Aerial photography, view from above. Overlap is shown in striped.

Aerial photography side view. Overlap is shown in striped.

Satellites

Satellites can make stereoscopic images in two ways:

A satellite takes one image sideways when it is located at orbit (i) around the Earth. When it is one orbit further (i+1) it makes another image of the same area.
A satellite takes an image forward. A few seconds later, it makes an image of the same area, backward.