by Hurbert Mara
Abstract:
Tens of thousands of fragments of ceramics (called sherds for short) are found at every archaeological excavation site and have to be documented for further archaeological research. The traditional documentation is based on the profile line, which is the intersection of the sherd along the axis of symmetry in the direction of the rotational axis. Traditionally this is done by experts by manually drawing the profile line, using different tools like a Profilkamm (profile comb), flexible wires, circle- templates, etc. to estimate the axis of rotation and the profile line. The traditional drawing is error prone and time consuming, therefore a semiautomatic method using a Profilograph was introduced to increase accuracy. Since the measurement is still manually, the time for drawing was not decreased. We propose a fully automatic system for the estimation of the rotational axis and the profile line. For data-acquisition we are using acquisition methods based on the principle of structured light, which have also been compared in respect to accuracy and performance to traditional methods of documentation. Based on continuous experiments and comparisons we show a new method for estimation of the rotational axis and the profile line, which is inspired by traditional archaeological methods. The methods shown in this thesis were tested on synthetic and real data. The experiments with real data were done at the archaeological excavation in Tel Dor in Israel. The results for estimation of the profile line and the comparison between the manual drawings, the Profilograph and the 3D- acquisition by structured light are shown in this thesis. Furthermore methodological experiments of geometrical surface analysis are shown, which demonstrate the possibility of estimation of ancient manufacturing techniques of ceramics. Finally an outlook towards detection of lines and analysis of painted ceramics is given.
Reference:
Documentation of Rotationally Symmetric Archaeological Finds by 3D Shape Estimatimation (Hurbert Mara), Technical report, PRIP, TU Wien, 2006.
Bibtex Entry:
@TechReport{TR103,
author = "Hurbert Mara",
title = "Documentation of Rotationally Symmetric
Archaeological Finds by 3D Shape Estimatimation",
institution = "PRIP, TU Wien",
number = "PRIP-TR-103",
year = "2006",
url = "https://www.prip.tuwien.ac.at/pripfiles/trs/tr103.pdf",
abstract = "Tens of thousands of fragments of ceramics (called
sherds for short) are found at every archaeological
excavation site and have to be documented for
further archaeological research. The traditional
documentation is based on the profile line, which is
the intersection of the sherd along the axis of
symmetry in the direction of the rotational
axis. Traditionally this is done by experts by
manually drawing the profile line, using different
tools like a Profilkamm (profile comb), flexible
wires, circle- templates, etc. to estimate the axis
of rotation and the profile line. The traditional
drawing is error prone and time consuming, therefore
a semiautomatic method using a Profilograph was
introduced to increase accuracy. Since the
measurement is still manually, the time for drawing
was not decreased. We propose a fully automatic
system for the estimation of the rotational axis and
the profile line. For data-acquisition we are using
acquisition methods based on the principle of
structured light, which have also been compared in
respect to accuracy and performance to traditional
methods of documentation. Based on continuous
experiments and comparisons we show a new method for
estimation of the rotational axis and the profile
line, which is inspired by traditional
archaeological methods. The methods shown in this
thesis were tested on synthetic and real data. The
experiments with real data were done at the
archaeological excavation in Tel Dor in Israel. The
results for estimation of the profile line and the
comparison between the manual drawings, the
Profilograph and the 3D- acquisition by structured
light are shown in this thesis. Furthermore
methodological experiments of geometrical surface
analysis are shown, which demonstrate the
possibility of estimation of ancient manufacturing
techniques of ceramics. Finally an outlook towards
detection of lines and analysis of painted ceramics
is given.",
}