A collaboration between opto-technicians at Darmstadt University of Applied Sciences and ecologists at Darmstadt Technical University has spawned results far beyond expectations. Working together they have developed a device that can help to preserve insect specimens which would otherwise be irredeemably lost. Moreover, its working method renders the Darmstadt insect scanner globally unique.

By Daniel Timme, 11.9.2018

Insects are crawling, scuttling and flying into the public gaze with increasing tenacity, as their huge significance to the planet’s well-being bores deeper into our collective consciousness. We read of how bees’ untiring busyness is utterly essential to Nature, and to our species as well, but that they are dying out; of an alarming reduction in biodiversity as a result of catastrophic levels of intensive farming, for should we fail to curb certain aspects of our behaviour patterns, we can be sure the generations to follow will never meet most of the many insect species that swarmed around us as children, landing on our own, outstretched fingers. However, a similar effect, yet with distinctly more scientific profundity, can now be produced using a technical device developed by a collaborative project between the Darmstadt University of Applied Sciences (h_da) and Darmstadt Technical University (TU Darmstadt). The ‘DISC3D’ (Darmstadt Insect Scanner 3D) focuses, quite literally, on insects, capturing every facet of them and enabling the Darmstadt Insect Scanner to play an important role in preserving knowledge of the various species.

A collaboration between opto-technicians at Darmstadt University of Applied Sciences and ecologists at Darmstadt Technical University has spawned results far beyond expectations. Working together they have developed a device that can help to preserve insect specimens which would otherwise be irredeemably lost. Moreover, its working method renders the Darmstadt insect scanner globally unique.

Insects are crawling, scuttling and flying into the public gaze with increasing tenacity, as their huge significance to the planet’s well-being bores deeper into our collective consciousness. We read of how bees’ untiring busyness is utterly essential to Nature, and to our species as well, but that they are dying out; of an alarming reduction in biodiversity as a result of catastrophic levels of intensive farming, for should we fail to curb certain aspects of our behaviour patterns, we can be sure the generations to follow will never meet most of the many insect species that swarmed around us as children, landing on our own, outstretched fingers. However, a similar effect, yet with distinctly more scientific profundity, can now be produced using a technical device developed by a collaborative project between the Darmstadt University of Applied Sciences (h_da) and Darmstadt Technical University (TU Darmstadt). The ‘DISC3D’ (Darmstadt Insect Scanner 3D) focuses, quite literally, on insects, capturing every facet of them and enabling the Darmstadt Insect Scanner to play an important role in preserving knowledge of the various species.

Graduate biologist Michael Heethoff is part of the ecological networks work group at TU Darmstadt, where one of the ecologist and entomologist’s fields is a morphological study of biodiversity. Morphology describes the scientific examination of a creature’s form, just as biodiversity describes the extent of biological variety in all living species. Heethoff and his colleagues had at their disposal instruments including state-of-the-art x-ray tomographs to help them study and catalogue the surviving global insect populations. Yet as the 45 year-old explains “to date there had been no simple, easy-to-use measurement procedure to measure the surface area and volume of an insect, for instance.” Such a tool would be an asset as both of these factors are important to understanding the specific physiology of a species. “Then a colleague who had studied at h_da suggested I get in touch with their opto-technicians” recalls Michael Heethoff. The first meeting occurred but a few weeks after Heethoff’s initial email to the h_da sent in September 2014. There, Heethoff and his colleagues Sebastian Schmelzle and Nico Blüthgen met with h_da professors Stephan Neser and Bernhard Ströbel from the mathematics and natural sciences faculties to discuss and assess possible approaches. The most promising appeared to be a suggestion from Stephan Neser: multi-image stereo photogrammetry, also known as ‘Structure from Motion’. “From our experience with optical 3D measurement techniques, this method seemed the most suitable” says Ströbel.

At 65 years of age and lecturing in his final semester in his specialist fields opto-technology and image processing at the time, physicist Bernhard Ströbel continued his interest in the project once he had officially retired. Impetus from Stephan Neser helped the project being managed by Heethoff and Ströbel to really take off. As part of their practical training, two students from the TU constructed the first provisional measuring station at the h_da towards the end of 2014. As Michael Heethoff says, the Darmstadt scientists all got on well together. “There aren’t that many natural overlaps between our fields, not regarding knowledge and vocabularies, but the fact that I’ve been interested in 3D imagery for quite a few years helped – it all went extremely well, working together.”

However, many of these attempts prove to be highly elaborate, or they only deliver results of a dubious quality. As Michael Heethoff explains, in cases where specimens were only photographed from two or three different perspectives, several of the characteristics specific to a species, and which need to be clearly apparent in order to clearly differentiate between similar species – such as the length of legs, the shape of jaws etc. – have still to be guessed at. By contrast, the Darmstadt insect scanner renders all of these facets clearly discernable, for its images reproduce the finest hairs, wing structures or true colours in full detail. They enable experts to rotate, flip and carefully inspect models of the insect specimen generated from the graphic data on the screen –  moreover, the digitalized information can easily be stored and shared all over the world, meaning the DISC3D could make a huge contribution and provide impulses to the world-wide transfer of knowledge on bio-diversity. Michael Heethoff believes this development will enable ecologists to do much more in the future, because for along time, research was almost limited to lists of species being compiled, whereas “in the meantime, we are focusing increasingly on the changing characteristics of ecological systems and the organisms they comprise of – with these insights being related to phenomena and influencing factors such as climate change and the ever-increasing use of land for intensive farming.” The three dimensional perspectives afforded by the insect scanner make even the smallest, finest features apparent, enabling how these are changing to be easily assessed. “We’re now able to link these variations to simultaneous changes in environmental circumstance and provide explanations for them.” Even evolutionary biology cannot deliver definitive assertions on concrete causation of such variations, “yet at least this method of observation enables us to clearly identify potential mechanisms – and not merely patterns.”

The method employed by the Darmstadt approach differs markedly from other attempts to digitalize insect specimens, by dint of two techniques being combined: an automated all-round imaging of the specimen and scanning it with highly expanded depth of field. “Our scanner provides a huge advantage over other techniques in digitalizing collections” says Michael Heethoff, “for one thing, we no longer need to pick the delicate specimens up to change their pinned position, plus, the scanning process is fully automated.” Thus reducing effort, time and outlay.

“We’re the only team able to routinely create 3D models of insect specimens” says Bernhard Ströbel. The process is as follows: the specimen is pinned and mounted between two hemispheres where it is illuminated indirectly. Then the specimen is rotated at regular intervals along two axes by electric step-motors. “ Each insect gets to pose around 400 times for the camera“, says Bernhard Ströbel. At each interval the camera is moved backwards and forwards on a macro slide to record the very precise image of every single aspect of the specimen. The resulting roughly 25,000 images are then calculated to match one-another. These images with their expanded depth of field and true colour and are then used to generate true-to-scale and colour, three-dimensional models of the specimen. “We just load the specimen, press a few keys to set the parameters required then let it scan away automatically – without even anyone having to be there”, explains Ströbel, which makes our scanner unique across the world, so far as we can tell.”

The amounts of data generated are considerable. As Bernhard Ströbel points out “if we were to store the raw imaging data then we’d have half a terabyte, so 500 gigabyte, from each scan.” To this end, the data is processed during the scanning in parallel to reduce the amount; using this combination of processes the amount of data for the final 3D model of a specimen can be reduced to around only 600 to 700 megabyte, on average. As Ströbel concedes “what we are still finding a nuisance are extremely fine body parts, such as hairs, bristles and some incredibly thin and transparent wing parts – rendering these into three-dimensional models is tricky, to say the least.” Additionally, some of the partially hidden body parts and details, of dark-coloured insects in particular, are very difficult to capture: “we’re working on both.”

Noting the times required for each scan Bernhard Ströbel says “in January we were still looking at up to ten hours per scan, now we’re down to five hours at most.” A rule of thumb dictates that smaller takes longer. As Ströbel explains “small depth of field plays an important role with smaller insects because they need to be enlarged the most.” Processing and calculating the images takes up the most time. “In essence it comes down to a question of processing capacity and software. It’s our aim to be able to complete a scan within one hour – and that with an improved resolution, if possible!”

Bernhard Ströbel says yet another exciting use for the data sets generated by the DISC3D would be 3D prints. “Just to try it out we printed a beetle at a scale of 10:1 and it looked pretty impressive, and larger models are naturally feasible as well.” One problem that persists with 3D prints is that the objects don’t appear in true colour – but specialists are already working on this. Ströbel can well imagine that the DISC3D used in connection with 3D printing could pay dividends in museums’ educational programmes. Models of creeping, crawling and flying creatures could serve to inspire young and old to join in.

The die has since been cast, following their recent publication in the zoological online journal ‘ZooKeys’, leaving the creators of the Darmstadt insect scanner to wait and see how both the scientific community and the public at large react. Perhaps it could even help to generate a welcome wave of attention – that at the end of the day even benefits the industrious insects to whom we owe so much.