Accuracy assessment of a compact stere camera for measuring marine fishes in the Atlantic Ocean off the southeastern US

Fishery-independent length-frequency distributions are essential for most stock assessments, but they can be challenging to obtain for many species. Using a popular compact underwater stereo video system, we calculated the difference between the stereo camera software measurements and the actual lengths (Difference from Actual, DFA) of three target sizes at six regions within the cameras’ field of view. We found that targets centered in the frame, perpendicular to the stereo unit, produced the most accurate measurements [mean |DFA| = 0.21 cm (SE 0.02)], with variability increasing as the target was near the side of the frame and angled [mean |DFA| = 0.95 cm (SE 0.07)]. Measurement accuracy worsened over the course of a field season, with stereo measurements more variable and less accurate after 129 deployments than immediately following calibration or even after 80 deployments. Verified field measurements varied little among frames within the same video and were accurate to 1‐cm 90% of the time. In short, the SeaGIS “Aluminum Base Bar System: Narrow Separation” stereo units can provide reasonable measurement estimates for stock assessment purposes under a variety of circumstances with regular calibration.