Currently aligning the view orientation to match a background photo is hard.  There are two current options: either approximate it by eye or use a fairly complex method involving projection lines and a dummy “backplate”.  Matching the 3D model to the background would help users create renderings and visuals of their model in a realistic setting - this has applications in architectural engineering and product design.

Proposal: a new camera type called "Background matched" in addition to "Aimed at target" and "Floating".  The user would be prompted to map some points (vertices / sketch points) from the 3D model to their corresponding positions on the background.  If three (or four?) points are defined it is possible to calculate the position of the camera in 3D space.  Camera view angle and aspect ratio can then be derived from the background image size.

An alternative proposal is to use a similar method to this functionality in Google Sketchup:

http://support.google.com/sketchup/bin/answer.py?hl=en&answer=94921