This contour model is remarkable for 2 reasons: 1 it is derived from KAP imagery and 2 it has captured in 3D some very subtle detail of the earthwork: the firing step and the ‘bridge’ bank across the ditch have been recorded. These are features which are hard to see on the ground. The contour interval is 5mm
I have been trying out various pairs of photos in AGIsoft Steroscan and have begun to find out what sort of image pairs it is happy working with. I have a good strip of shots of the Bulwark to work with and I have found quite a few just don’t work. This contour model is derived from this pair:
The offset is something like 20% and this is sufficient for a surface model to generate with out failure or ‘stepping’. This is the auto alignment of the pair:
Other models I have attempted failed because of inadequate offset between the images. I have relied on burst mode on the camera and kite wobble to get the base but this is not enough. The offset should not be left to chance movements of the kite ( although it’s a nice surprise when it is!). By walking the kite 5 -6m from shot to shot (or burst to burst in my case) and pointing the camera at the required area the offset worked. In this example the height of the camera was 25m AGL .
The disruption to the DTM ( trees etc) was edited out in Meshlab before transfer to CAD for contouring with TheoContour.
The tilt of the image pair was taken care of by Steroscan although the out-put surface was tilted in 3D space and had to be re-aligned to ‘make sense’ in CAD.
More on TheoContour here.
More on StereoScan here.
More on Meshlab here