Jonathan A Brecko, Aurore Mathys, Wouter Dekoninck, Maurice Leponce, Didier Vanden Spiegel, and Patrick Semal (2014)
Image stacking: a semi-automated approach allowing high quality mass digitization
In: Digital Specimen 2014, Berlin.
Scientific institutions like museums maintain large collections enabling present and future research. Thanks to the digitization of the collections, in most of these institutes, it enables researchers across the globe to see which collections might be interesting for their projects. However as most of these databases only provide descriptive information and/or metadata, it remains impossible to study these digitized specimens from a distance. As the most precious parts of the collections, like type specimens, are the most requested ones by fellow researcher, these become, inevitably, the most handled ones. Unless a policy exists not to handle them at all, which unfortunately, prevents research. To make sure that collection material like type specimens remain documented for future research and can be studied trough the internet, digitization is the key, 3D or 2D. The only challenge is to chose the right digitization method for the right material and or size (Mathys et al., 2013). Especially small specimens which are often found in insect and invertebrate collections tend to be difficult to digitize in 3D as fine structures can only be seen in µCT recordings, which are still quite expensive pieces of equipment. However, 2D image might provide enough information to conduct for instance taxonomic research. Image stacking is the only way to capture enough detail in a single picture as the low depth of field of camera lenses, makes it almost impossible to get the complete object in focus, unless the aperture is stepped down. However this results in other aberrations as the optical resolution reduces due to the diffraction effect. Thanks to the large computational power of todayʼs workstations, it is possible to do image stacking rather easily. The only remaining challenge is taking the individual pictures. Readily made commercial stacking columns do exist, but are too expensive to provide one to each department or research group. In this paper we will present the results of a low cost approach consisting of a DSLR camera attached to an automated macro rail with a custom build light tent. As there are no over- or underexposed parts on the resulting images, they are good enough to allow publication without the use of a post-processing software. More importantly, during the photo-shoot of one specimen another can be prepared, alcohol based or dry, for the next shoot. When similar specimens are digitized at the same time, light and aperture settings stay more or less the same, providing a fast and smooth workflow. The stacking of the images, done in Zerene Stacker, can easily be started as a batch process at the end of the working day or during the night. As the total package of this system can be purchased for around € 2.5k several digitization centers can be started in different department allowing faster digitization of the type material.
Document Actions