Photogrammetry- camera position

[vrivanov92]

Hello,

I'm new to photogrammetry and i want to ask if knowing the camera position will help model construction?

I want to scan a small object (5cm x 5cm x 5cm) and the camera can be moving on spherical plane.

Will knowing the camera position enchance the 3D model or build it faster?

[Taylor]

It depends on the software you're using to construct the 3D model and what your project requirements are for establishing ground control and an absolute coordinate system.  For example, this paper compares the accuracy of various methods of recording the exterior orientation parameters of the camera:  http://homepages.lboro.ac.uk/~cvrw5/papers/Kirchhoefer.pdf

 

However, up-to-date Structure from Motion (SfM) software (e.g., Photoscan) can usually generate a good, reasonably accurate model without knowing camera positions or orientation in advance, if you include an appropriate target scale in the images and follow guidelines for good image capture as described on CHI's photogrammetry page and in documents you can find in the sidebar there.  Knowing how to optimize the sparse point cloud (as taught in CHI's classes and in other sources and tutorials you can find) will help to generate a more accurate model with less noise.  I'm not sure what you mean by moving the camera in a spherical plane, but if you follow the guidelines for capturing close-range photogrammetry images, you should be able to construct a good model using any of several available software packages.  Best of luck!

[cdschroer]

We have updated our photogrammetry technology page with a bit more info about structure from motion approaches to photogrammetry processing, and also some tips for camera settings and figuring out your sensor network.

 

Check it out: http://culturalheritageimaging.org/Technologies/Photogrammetry/