We argue that for a method to be useful in practice, it is necessary t
o perform its sensitivity analysis and to investigate its robustness t
o measurement errors. We present here a complete sensitivity analysis
of the 3D reconstruction method based on projective geometry. We use t
his sensitivity analysis to best design a system for the inference of
the shape of a block of granite from cameras placed at 90 degrees angu
lar separation. The system has been tested on both real and synthetic
data.