Procedurally constructing a building from an image

Question

I have a large data set of images. These images contain buildings. I am looking for an intelligent way to extract data from these images.

Input

The program needs to be able to extract the image of the face of the building.
Correct the image so it is "front facing" (some of the images are taken at angles to the face of the building)
Assign tags to detail elements of the buildings face (windows, doors etc) and give these elements coordinates within the face of the building.
Assign tags to textures of the building (ie Brick, concrete, misc 1, misc 2)
Group similar building images together (using size and shape)
Group similar detail elements and texture elements together into subdirectories of images.
Assign coordinates (which are provided with source) to each building image

a clean database of all building face imagery, with associated coordinates, textures, and detail elements with image coordinates, easily indexable via Javascript or PHP.

I appreciate any and all advice on this project. Any links to useful code snippets/libraries/projects.

score 0 · Answer 1 · answered Apr 13 '12 at 14:40

I think it's more likely that you'll get this done using Mechanical Turk than with a program

If so, your job would be to make a very simple to use GUI for a person to provide information you need to do the transformations and extractions.

Collect 4 points that are the corners of the front face
Extract that quadrilateral from the image
Use a quadrilateral warp to turn that into a rectangle
Show the user this image
Let them tag the parts
Let them fill in meta-data (Concrete, etc)
Assign a QA task to a different person that just checks it over (or alternatively, give the same task to a different person and QA if they disagree)