Hi,
My goal is to estimate volume of the log piles in a mill yard. I had acquired 137 images (with 80% overlap at 40m above ground) from a UAV Platform - all in jpeg. I am trying to build 3D point clouds (and eventually DSMs) using Photoscan pro. I have negative elevation values (see attached screen shot ). I am new to this environment and had been following the workflow shown in the tutorials.
1. No markers used but have image centre GPS co-ords from the sensor. I used sensor altitude agl and not gps altitude. Projection WGS84.
2. Aligned the photos using high accuracy and "reference" pair-selection. Optimised cameras after removing points with high reprojection error (over 1 pixel).
3. Built high quality dense point clouds (as in the image), Mesh with height surface type, decimated the mesh, filled holes to build texture (adaptive ortho mapping mode and mosaic blending).
Unable to figure out why I have negative elevation. Any insight to where I am going wrong is much appreciated?
Thanks!