1. INTRODUCTION
The methodology of terrestrial photogrammetry changed
significantly in the last years thanks to the improvement of
matching algorithms and to the introduction of simplified digital
tools that easily permit the 3D object model reconstruction
without stereoscopic skills.
In this work an experiment has been conducted on a real case,
aiming to verify what accuracy may be reached at close-range
photogrammetry when a semi-metric and a non-metric camera
are in use, applying two strategies for image processing and two
strategies to solve the orientation problems.
Object of the test is the Rolandino dei Passaggieri Tomb located
in the famous San Domenico square in Bologna (Italy), adopted
for its elevated historic-architectural meaning, for its ideal
position to traditional photogrammetric survey realization
purpose and for its dimension (length, height, width),
representative of architectural and archaeological applications.
In the frame of this work, the following activities were realized:
• a topographic survey was performed to define the 3D
coordinates of natural Control Points using a modern
prism less Total Station (Topcon GPT 6001);
• direct object distance measurements were collected and
accurate monographies were redacted;
• pseudo-normal, and horizontal, vertical or oblique
convergent photos of the tomb with different scale
factors were acquired by means of Leica R5 semi-metric
camera and a Nikon Coolpix 5400 non-metric digital
camera;
• a camera calibration work was performed using
PhotoModeler v5.2 (EOS Systems Inc.) calibrator
software tool in order to obtain the parameters of the
inner orientation together with full lens distortion on the
non-metric camera;
• images processing (orientation parameter estimation and
restitution) using PhotoModeler’s monoscopic system
was realized by means of Control Points (method 1) and
using Distance Constraints (method 2);
• images processing was even realized by means of a
Digital Photogrammetric Workstation (SOCET SET
v5.2, Bae Systems), that permits manual and automatic
stereoscopic plotting;
• vector and raster outputs were produced by means of the
monoscopic system and the DPW;
• a virtual 3D model was generated in the format VRML.
At the end, the different solutions were compared in order to
evaluate the accuracy and the reliability in restitution of the
object shape.