Photogrammetry is a technique to extract geometric information from two-dimensional images to take measurements of photographs. It is used for recovering the (spatial) positions of surface points of overlapping photos from all perspective angles to create a digital reconstruction of spaces or objects (3D) based on photographic (2D) data. The rendering process turns the so created 3D spaces to images once again.
The way how photogrammetry was used here is at the edges of its normal use conditions: in this case it was not possible to shoot photos from all the possible angles:
As being fence spikes the very function of these objects is exactly to hinder the access to the other side.
Thus, the so created digital objects and spaces were condemned to fail:
They became glitches, the 3D spaces turned into zones that are filled with all kinds of distortions: some of the surfaces collided, some details went completely lost, shadows were interpreted as forms, shapes as holes in the landscape or as mere image projections.
They became a very indeterminate mix of depths and flats.