Quote:
Originally Posted by jyl
So I take it the arch pictured above would be relatively easier to model using photogrammetry than, say, the statue was?
Less work using Zephyr than to measure and manually build a CAD model? I could build a pretty accurate gross-scale Sketchup model of one of the arches in a few hours, but manually modeling the texture and detail would be hard.
I'm pretty excited about this . . .
People in my neighborhood would be thrilled to get little models of either as a token of appreciation. In fact, some would pay money for larger models, maybe enough for a fundraiser . . .
|
Photogrammetry uses multiple pictures of an object and tracks reference points.
Imagine rotating 360 around an object. By acquiring and tracking common reference points you can determine the shape of objects as the spatial relationship of those points change respective to each other.
From that you generate a point cloud that gets converted to your mesh.
The beauty is, it also stitches the images into a texture map for you.
Fun stuff.
As far as photogrammetry is concerned both objects would be the same basic amount of work for you. Far less than trying to build it yourself.
The tricky part will be getting the overhead shots on such a tall object.
Grab Zephyr and a camera.
You should be able to generate a texture mapped model of anything you have laying around the house in an hour first time out.
Just avoid glass, transparent objects. For those we spray paint them with a matte primer.
Reason being, transparent objects don't have consistent reference points cause, well, they are transparent.