Page 194 - Green - Maritime Archaeology: A Technical Handbook. 2nd ed

Chapter 6: Photogrammetric Techniques 173
vertical, one of which is using the grid frame as discussed earlier. An alter- native is to mount a small bubble level on the camera bar to help in level- ing to horizontal, but this can be an additional complication because the photographer will already be trying to adjust the longitudinal and lateral position and the height and the camera format orientation, all of which require considerable concentration through the viewfinder. Because it is distracting to have to glance away from the viewfinder in order to observe the small bubble level, a solution is to mount the bubble level so that it appears in the corner of the the viewfinder’s field of view, utilizing the great depth of field of the level.
A most practical approach to this form of survey control is to mount the camera on a photographic tower (Figure 6.4). There are considerable advantages to using a photo tower. First, it reduces darkroom time. Next, the orientation of the camera is set vertical to the grid frame and the dis- tance is always the same. Thus, once the enlarger is set up, all photographs can be run off without further adjustment. The tower can either be mechan- ically leveled or set up each time in a random orientation. This advantage is offset by the inconvenience of maneuvering the tower under water. This technique will be discussed in detail in Section VI. Another alternative, where the practicalities of a large cumbersome four-legged tower may be difficult, is to use a bipod frame or stand (Figure 6.5).
D. NETWORK CONTROL
Atkinson et al. (1989) recently developed a system for producing accu- rate photomosaics as a result of the least-squares adjustment program which was designed by Atkinson et al. (1988) for site survey work (discussed in Chapter 4, Section VI.C). This system was first tried during a survey of the extensive hull structure of the Ko Si Chang 2 wreck site in the Gulf of Thailand in 1987. It involved making a photomosaic of the site and taking a number of measurements with a tape measure between selected survey points on the site. These points were then used in a least-squares adjust- ment program to calculate the coordinates (see Atkinson et al., 1988). The output of the program provides calculated X and Y coordinates for each station together with residuals for each measurement.
However, there are now a number of more sophisticated and more user- friendly programs available, and although the basic principals are the same, the least-squares program is obsolete. Once the coordinates have been located they can then be used to scale the photomosaic of the site. As the tags used in the tape survey can be identified on the photomosaic, it is therefore possible to determine the scaling factor. Usually it is found that the scale of a photomosaic is marginally smaller than that indicated by the