Modeling for architecture: Photo matching in Blender
The most common starting point for architectural visualization projects is a CAD file with some kind of technical drawing of the visualization subject, or at least a few sketches. With a CAD file we will import the lines and work with them to transform 2d lines into a 3d model. But, sometimes the starting point of a modeling project will be something like a photo. This is a usual way to create facelift projects for old buildings, where the facade should be represented to add or change elements.
In Blender we have a few scripts that can make the camera match easily, but with only a few changes on the camera and a 3d mesh cube, we can already try to match a photo with a 3d model.
The tutorial below shows exactly how to do that with a building:
Camera matching in Blender tutorial from mookie on Vimeo.
When the 3d cube of Blender is perfectly matched with the photo, the artist starts to add details and edges to create the windows and other facade elements. By the end, it will be quite easy to recreate the same facade with new details or changes for a face lift.
If you have a photo of a real building that must be modeled with Blender, the technique showed on the video will help to achieve a photo match quickly.
Brilliant.
Don’t you have to adjust the lens definitions in blender to match the ones used in the picture ?
But, what if we don`t have this information?
In this case, the adjustments will have to be based on experimentation anyway.
Regards
Maybe a tip about exif is in order: http://en.wikipedia.org/wiki/Exchangeable_image_file_format
From the exif we can get the focal point, size of lens etc.
you can get the information in sites like flickr and different desktop methods