International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 5. Hakodate 1998 
DESIGN AND CALIBRATION OF AN UNDERWATER STEREO-VIDEO SYSTEM 
FOR THE MONITORING OF MARINE FAUNA POPULATIONS 
Mark R. Shortis 
Department of Geomatics 
University of Melbourne 
Parkville 3052, AUSTRALIA 
M.Shortis@unimelb.edu.au 
Euan S. Harvey 
Department of Marine Science 
University of Otago 
Dunedin, NEW ZEALAND 
Euan.Harvey@stonebow.otago.ac.nz 
Commission V, Working Group IC V/III 
KEY WORDS : Video camcorders, underwater, calibration, stereo matching, marine fauna, environmental monitoring 
ABSTRACT 
Assessment of age and size structure of marine populations is often used to detect and determine the effect of natural and 
anthropogenic factors, such as commercial fishing, upon marine communities. À primary tool in the characterisation of population 
structure is the distribution of the lengths or biomass of a large sample of individual specimens of a particular species. Rather than 
use relatively unreliable visual estimates by divers, an underwater stereo-video system has been developed to improve the accuracy 
of the measurement of lengths of highly indicative species such as reef fish. In common with any system used for accurate 
measurements, the design and calibration of the underwater stereo-video system are of paramount importance to realise the maximum 
possible accuracy from the system. Aspects of the design of the system, the calibration procedure and algorithm, the determination 
of the relative orientation of the two cameras, stereo-measurement and stereo-matching, and the tracking of individual specimens are 
discussed. Also addressed is the stability of the calibrations and relative orientation of the cameras during dives to capture video 
sequences of marine life. 
1. INTRODUCTION 
A main theme of global ecosystem monitoring programs is how 
the marine ecosystems of our planet will be affected by 
environmental impacts and how, in turn, this will effect global 
climate change. To this end, major research is required on the 
response of the marine fauna and flora to changes in physical 
and biological factors. With many reef fish occupying positions 
at or near the top of food webs, their abundance and population 
structure are highly dependent on the availability of food and 
the state of their environment. Accurate information on the size 
structure of a fish population, when linked with knowledge of 
the biology of the species, can allow analysis regarding fishing 
intensity, environmental impacts and rates of recovery 
(McCormick and Choat, 1987). Species of reef fishes are 
therefore useful indicators of the status of near shore temperate 
and tropical ecosystems. Visual census techniques (Thresher 
and Gunn, 1986) utilise SCUBA divers to count reef fish 
abundance and in some cases estimate the lengths of reef fish to 
determine the size frequency or mean length of a population. 
These visual census techniques are used in marine reserves and 
sanctuaries around the world to monitor whether changes in the 
abundance and size frequency of species populations are 
occurring (Francour, 1994). More recently they have been used 
in fisheries management as a tool for assessing the standing 
stock or biomass of individual species based on the relationship 
between the estimated length and the weight of individual fish 
of a certain species (Russ and Alcala, 1996). 
However, making accurate and precise visual estimates of the 
length of objects underwater is extremely difficult and requires 
the observers to be well trained and experienced (English et al, 
1994). The estimation of the length of an object underwater is 
complicated by the refractive effects of water, which increases 
the apparent size of objects and causes objects to appear to be 
closer to the observer than the actual range. Further, 
researchers using SCUBA are not efficient workers when 
performance underwater is compared to similar activities in the 
air (Hollien and Rothman, 1975). In addition, the sampling bias 
and errors resulting from the detrimental physiological effects 
related to SCUBA diving can be significant (Baddeley, 1965) 
792 
and where data from visual size or length estimates have been 
published, few authors include the precision or accuracy of the 
data. Problems with long-term studies occur when different 
observers may be involved in making estimates of size or length 
of marine organisms at different spatial and temporal scales. 
Even though calibration procedures are used by some 
researchers (Bell et a/., 1985) inter-observer variability has the 
potential to cause major biases. If the data collected is to be 
used to compare the size estimates recorded for different times, 
places or species then it is important that the level of precision 
and accuracy is known to enable rigorous analysis of the 
comparisons. Due to observer error and biases it is probable 
that many studies lack the statistical power to detect small 
changes in the length of the organisms being studied (English et 
al., 1994; Fairweather, 1991). To overcome the problem of 
subjectivity in visual estimates and enhance accuracy and 
precision, an impersonal system of measurement is preferable. 
Clearly, any impersonal system of measurement must be 
technology based, but within the limits imposed by the 
underwater environment and finite resources of research 
organisations. 
Many marine scientists and biologists have experimented with 
conventional and video imagery. For example, Klimley and 
Brown (1983) describe the use of stereophotography for 
estimating the size and dispersion of free swimming sharks. The 
system was viable underwater, convenient to use for 
measurement and could be developed or purchased at a 
reasonable cost. As a consequence, stereo-video cameras were 
quickly adopted for a wide range of applications in the marine 
environment (Hamner et al., 1987; Vrana and Schwartz, 1989). 
In recent times there have been rapid technological 
improvements in video cameras which has improved the utility 
and accuracy of such systems. 
Metric photogrammetry has been used specifically for various 
types of biological recording and analysis, generally using 
stereo photographs. Film-based stereophotogrammetry has 
been used to make many types of biological measurements, such 
as the demography of underwater plants (Kaczynski and 
Szmeja, 1988). Conventional film and video systems have been 
  
used for 
rig inspe 
(Turner, 
used to r 
al., 1996 
This pap: 
system S| 
of marin 
The long 
system v 
and shap 
the syst 
algorithn 
basic har 
video te 
cameras. 
on enabl 
and mea: 
by marin 
The desi 
factors, | 
estimate: 
A typic: 
metres, 
sampling 
affecting 
the cam 
camera 
manipul 
the over 
base se 
manoeu 
these tw 
metres, 
rage of 
The ster 
mountec 
buoyanc 
video c: 
resolutic 
6.3 mici 
air is ap 
As the 
maximu 
The foc 
water) t 
metres. 
optimis: 
these d 
competi 
increase 
precisio 
the mea 
of view 
The qu 
notorioi 
priori i 
been bc 
and the 
object s 
and de 
confide 
poorer : 
to deptl