Page 177 - Green - Maritime Archaeology: A Technical Handbook. 2nd ed

156 Maritime Archaeology: A Technical Handbook, Second Edition
be remembered that PhotoModeler requires discrete targets or points that can be clearly identified in a variety of photographs. Thus, objects that are smooth and featureless or “fuzzy” do not work well with this system and the photo tower experiment hides this problem. This is discussed in Green et al. (2002). HPASS performed remarkably well, with an SD = 68% of underwater interpoint hand measurements but with a recording time of 20 minutes. However, this must be offset with the time to set up the control transponders (about 30 minutes). Obviously this situation would improve as the number of points increased because, like the hand-measured control points, the system becomes more efficient with a large number of mea-
Figure 4.30 A perspective view of the photo tower obtained by exporting from PhotoMod- eler to the three-dimensional graphics package Rhino. This image is a “screen grab” and thus low quality, although this image can be utilized at high resolution. Note the dimensions are to scale.