The answer to this question will depend on the Structure from Motion software being used. The two primary components are:
- The ability to use high-accuracy geotags.
- Loki provides highly accurate geotagged images with accuracy values written to the XMP metadata. We also produce a simple .csv text file for applications unable to read that metadata. PhotoScan and Pix4D both read the positions and accuracy data from the .jpg images directly.
- The ability to incorporate a camera calibration based on Brown’s Distortion Model.
- Necessary parameters:
- Focal length
- Principle Point X and Y
- Radial distortion coefficients 1,2, and 3.
- Tangential distortion coefficients 1 and 2.
Read the complete article here.