Portal de Periódicos da CAPES

Underwater Photogrammetry for Archaeology

2012; Linguagem: Inglês

10.5772/33999

Pierre Drap,

Robotics and Sensor-Based Localization

Archaeological excavations are often irreversibly destructive, so it is important to accompany them with detailed documentation reflecting the accumulated knowledge of the excavation site.This documentation is usually iconographic and textual.Graphical representations of archaeological sites such as drawings, sketches, watercolors, photographs, topography, and photogrammetry are indispensable for such documentation and are an intrinsic part of an archaeological survey.However, as pointed out by Olivier Buchsenschutz in the introduction to the symposium Images and archaeological surveys, in Arles, France, in 2007 (Buchsenschutz, 2007, Introduction page 5), even a very precise drawing only retains certain observations that support a demonstration, just as a speech retains only some arguments, but this choice is not usually explicit.This somewhat lays the foundation of this work: a survey is both a metrics document and an interpretation of the site by archaeologist.The survey is a very important component of this documentation and its importance is largely due to the fact that the concepts employed by archaeologists during an excavation are closely related to space.The structure of the excavation is based on the concept of stratigraphic units.Inherited from geology and then formalized for archaeology by E.-C.Harris (Harris, 1979), stratigraphic units are linked by geometric, topological and temporal relationships.They are fundamental for the interpretation of the archaeological excavation.Two families of objects have to be surveyed: first, the artifact that we seek to position in space and of which we have a good a priori knowledge; and second, the area being excavated (in this case, a part of the seabed), often represented as a digital terrain model (DTM).Throughout this work we deal with these two aspects, artifacts and unstructured land, by addressing two different approaches; one using a priori knowledge through measurements and the second based solely on geometry.The first approach, based on the a priori knowledge that we have about the measured artifact, uses our knowledge of the object to compute its size and position in space.This method can also reduce the time required for measurements.The second approach, used to survey land for example, uses automatic tools coming from photogrammetry to compute a dense cloud of 3D points.www.intechopen.