Within the science of fish swimming biomechanics, high-speed video imaging is shifting from 2D to 3D recording using multiple synchronized cameras, a trend being driven by software advances in automated analysis.

This move towards 3D was recently demonstrated in a study ("Three-dimensional analysis of the fast-start escape response of the least killifish, Heterandria formosa") published in the Journal of Experimental Biology, in which three Mikroton EoSens CL MC1362 cameras captured high-speed movements of adult fish. Specifically, research findings highlighted the importance of 3D imaging in the analysis of fast-start maneuvers of adult fish in uncovering the versatility of their escape repertoire. Experiments were performed with adult female Heterandria Formosa (least killifish), a species of the live-bearing fish family Poeciliidae that naturally populates a diversity of habitats in the southeastern USA, including small freshwater lakes, streams and marshes.

Until now, fast-start escape responses of adult fish have only been studied in a horizontal 2D plane. In natural situations, however, predators may approach a prey from any orientation in the 3D space. For example, birds strike from above, while other sealife attack mostly from below or strike in a mostly horizontal plane. To evade predator strikes, fish make C-starts consisting of three stages in which the fish rapidly bends into a C-shape, bends in the opposite direction; and immediately straightens. Irrespective of the initial orientation, fish can escape in any direction to prevent being captured from above, below or straight on. Using automated 3D tracking of high-speed video sequences made possible with the Mikrotron cameras, the researchers revealed that fast-starts often contain a significant vertical velocity component, with large changes in pitch and roll angles only seen in 3D images.

Escape responses were recorded in a 9 × 9 × 12.0 inch (L x W x H) aquarium with a water level at 9 inches, giving the fish ample room to move. The central area of the aquarium was filmed using the three Mikroton EoSens CL MC1362 high-speed video cameras, set at 1040 × 1020 pixels resolution, 470 frames per second, and 1/1000 second shutter speed. Each camera was equipped with Voigtländer Ultron F=40 mm 1:2 aspherical compact pancake lenses, an Epix PIXCI E8 frame grabber, and was calibrated with direct linear transformation (DLT) by indicating the position of 72 points on a custom-designed 3D-printed frame. A Quantum Composers 9214 digital delay pulse generator synchronized the cameras.

To elicit an escape response, a 3.6-ounce weight was dropped from the top of the aquarium using a manual electric switch when a fish was in the field of view, preferably in a still, steady and straight position. The aquarium was lit uniformly from the top using an LED panel. Similar panels on the side switched on when the stimulus was dropped, with a delay of a few milliseconds. Between stimuli, the fish were allowed a rest period of at least 5 minutes. A maximum of five escape responses were recorded per fish per day, and a total of 437 escape response sequences were recorded.

Motion and deformation of the fish during its "escape" were reconstructed in 3D from the Mikrotron video using MATLAB software. An in silico representation of the fish and experimental set-up identified the best possible fit of a fish model to the recorded video frames. This model consisted of a tessellated 3D surface of the fish, with a specified position, orientation and body curvature. The experimental set-up was then virtually recreated by calibrating the cameras with DLT. An image of the model fish was projected onto the virtual cameras, showing how the high-speed video frames would look for a given set of parameters, such as position, orientation and curvature. Overlap between the projected image and the actual high-speed video frame gives an indication of the goodness-of-fit for this set of parameters.

Fish biomechanics consist of complex 3D maneuvers in both larval and adult fish, indicating that the nature of this motion pattern might have been oversimplified in previous 2D studies. Using the Mikroton EoSens CL MC1362 cameras, researchers established that 3D imaging could aid the understanding of all aspects of fish swimming biomechanics.