at approximately human interocular distance between centers of 

 lenses (70 mm) to produce the left and right video images that 

 are recorded on separate left and right VHS recorders. The 

 advantages of this system for shallow water work are that the 

 small, relatively inexpensive cameras and recorders can be 

 packaged in one portable underwater housing and that one does not 

 suffer a major financial loss if the housing floods. This 

 shallow water system produces excellent 3-D images that can be 

 viewed comfortably for long periods of time. 



For deeper water investigations that employ untethered 

 manned submersibles we use a system marketed by Stereographies 

 Corporation and designed in part by Lipton (1982). Two video 

 cameras are mounted parallel to each other in separate housings 

 outside the submarine and the images ported inside to a single 

 tape recorder onto which the video signals are multiplexed. 

 The images are presented alternately on a single monitor within 

 the submarine via stereo image alternation using electro-optical 

 shutters at the screen surface and on the glasses worn by the 

 video operator in the submarine, thereby separating the left and 

 right images and creating stereopsis. The advantage of this 

 system for use in submersibles is that the operator can view the 

 images as they are recorded and thereby insure that the best 

 possible information is collected during the dive. The single 

 monitor and recorder are compact, which is an important 

 consideration inside a submarine. The disadvantage of the single 

 monitor-recorder solution is that the equipment necessary to 

 multiplex two video signal electronically onto one video frame is 

 expensive. This system also produces excellent 3-D images that 

 can be viewed for long periods of time without eyestrain. Both 

 systems have sufficient resolution to permit subsequent 

 stereophotogrammetric reconstruction and measurement. 



3-D calibration of cameras 



Potel and Wassersug (1981) stated that "...the real power 

 of the generalized 3-D analysis will come in describing 

 ...geometry ...in natural habitats. A set of stable anchor lines 

 with markers (or fixed landmarks, such as coral heads) and two 

 cameras are the essential equipment necessary to collect data. 

 The Cameras are positioned anywhere. All that the investigator 

 needs to know is the 3-D positions of six reference markers on 

 the anchor lines. If the reference markers can be seen in each 

 image, it is even possible to film the scene with non-stationary 

 movie cameras, subsequently calibrating the perspective views in 

 each frame. " 



This protocol certainly would permit 3-D photography in the 

 field, but at sea, placement of fixed anchor lines is not 

 possible. We have proceeded instead using the approach of 

 Klimley and Brown (1983), who calibrated their 3-D still cameras 

 prior to use, thereby obviating the need for fixed reference 



310 



