Automated Charting of the Visual Space of Housefly Compound Eyes

The system and the protocol are designed to analyze the physio system of flies by mapping with minimal human intervention. With the system and the protocol, we can precisely know how the visual space of a fly size is organized. The system's advantages are the reproducibility and the speed of the mapping.

The study of compound eyes is an important part of animal vision research and has inspired several technical innovations which have produced artificial eyes. We have set an example on how to build and test an automatic device for scanning compound eyes. Details of algorithm development that bring the parts together need particular attention.

Start with collecting a fly from the laboratory-reared population. Prepare a restraining tube by cutting six millimeter from the upper part so that the tube has an external diameter of four millimeters and an internal diameter of 2.5 millimeters in the upper part. Place the fly inside the cut tube and seal the tube with cotton to prevent damaging the fly.

Then push the fly such that the head protrudes from the tube and the body is restrained in the tube. Use beeswax to immobilize the head while the eyes remain uncovered. Once done, cut the tube to achieve a length of 10 millimeters.

Then place the plastic tube containing the fly in the brass holder with one eye of the fly pointing upward and the holder resting on a tabletop. Adjust the orientation of the tube on the microscope as described in the manuscript to scan the whole eye within the range of the azimuth and elevation allowed by the setup. Set up the microscope by mounting an alignment pin on the azimuth rotation stage so that the X-Y position of the tip can be adjusted to coincide with the azimuth axis on the motorized stage.

While viewing with the microscope equipped with a 5X objective, use the Z-axis joystick to focus on the tip. Next, align the X-Y adjustment of the azimuth axis with the microscope's optical axis and use the X and Y axis joysticks to ensure that the elevation and azimuth rotary axes are pre-aligned with the centered pin. Manipulate the azimuth and elevation joysticks to check whether the pin is centered with respect to both degrees of freedom.

When well centered, the pin tip remains in the same position during azimuth and elevation rotations. Align and mount the fly with the elevation stage at zero degrees and holder on the azimuth stage. Then observe the fly's eye with the microscope.

After turning the illumination LED on, adjust the horizontal position of the fly to align the center of the pseudopupil. Then change the vertical position of the pseudopupil by using the rotating screw of the holder so that the deep pseudopupil is brought into focus at the level of the elevation axis. Next, line the deep pseudopupil with respect to the azimuth and elevation axes by centering it in the field of view.

When the setup is ready, switch the view to the digital camera mounted at the microscope and run the software initialization of the grace system, which includes initializing the motor controllers and the Arduino LED controller. To do so, open MATLAB version 2020a or higher version and run the MATLAB script. On the computer screen, confirm that the fly's pseudopupil at the center of the projected image.

Use the Z-axis joystick to bring the focus to the level of the corneal pseudopupil. Once the focus is aligned, run the auto focusing algorithm to obtain a sharp image at the cornea level. Then return the focus to the deep pseudopupil level by adjusting the motorized Z-axis stage.

Store the distance between the deep pseudopupil and corneal pseudopupil. Next, fine tune the pseudopupil centering with the auto centering algorithm followed by bringing the focus back to the corneal pseudopupil level. Rerun the auto focusing algorithm and zero the motorized stages at their current positions.

While scanning the eye, run the scanning algorithm to sample the eye images along the trajectories in five degree steps while performing the auto centering and auto focusing algorithms. After the sampling, turn off the LED and motor controllers. Later, process the images by applying the image processing algorithms.

In the study of optics of the fly eye, the image at the eye surface level shows the facet reflections and the pigment granule reflection in the activated state. The image taken at the level of the center of eye curvature illustrated the reflection of the arrangement of the photoreceptor cells in a trapezoidal pattern with their distal ends positioned at about the focal plane of the facet lenses. Two successive images were correlated to determine a shift in the translation of the facet pattern.

An image taken during a scan across the eye is shown with the facet centroids. After an azimuthal rotation of five degrees, the subsequent image is illustrated here. The centroid procedure could not identify all the facets.

A low local reflectance caused by minor surface irregularities, or specs of dust, resulted in erroneous centroids. The error was resolved by calculating a fast Fourier transform. The first ring of harmonics defines three orientations indicated by the blue, red, and green lines.

The inverse transformation of the harmonics along the three orientations yielded the gray bands. The right eye of a Housefly was scanned from the frontal side to the lateral side in 24 steps. The image shows the assembly of the facets as a Voronoi diagram.

At the start of the scanning, special attention should be given to the adjustment of the deep pseudopupil of the fly eye at the rotation center of the goniometric system. Here we apply epi-illumination microscopy. This method can be straightforwardly extended to fluorescence microscopy to study insects that do not have a reflecting pseudopupil.

Quantitative knowledge of the distribution of the visual axes of an eye will allow understanding of how physio systems are optimized for certain tasks such as hunting, mating, or detecting predators.