Imaging Considerations
____________________________________________________

Capturing images of types presents many challenges if the end results are to be scientifically valuable, virtual representations of physical specimens. A primary reason for undertaking such an effort is to enable scientists in remote locations to examine types without the need to handle them, with the ability to compare high-quality visual representations. This will only be worthwhile if the images carry sufficient information for the remote taxonomist to be able to state that the specimen in hand being compared with the type image is NOT of the same species. The effort will be of even higher value if the images allow for positive identification. If neither criterion is met, the imaging effort has little value.

In an attempt to present the best possible virtual access to types, two basic principles should be followed:

1. Always consult specialists for the taxonomic groups being imaged to determine which characters need to be documented in order for the images to be scientifically informative.


2. The images produced should faithfully represent the specimen as closely as possible.

The size and type of specimen will determine the most appropriate imaging system. In generally, with decreasing specimen size, problems with a lack of depth of field increase. This is particularly true when imaging subjects such as tiny insects using a microscope system. When specimens are very small (3 mm and less), optical limitations of equipment preclude obtaining an image with depth of field that spans the entire specimen in a single shot. The process of imaging insect specimens and other small groups is very different from capturing image data for other biological specimens, such as herbarium sheets, which are virtually two-dimensional and require simple scanning, or vertebrate specimens that are sufficiently large to reduce or eliminate the depth of field problem.

Imaging Systems

The following are imaging systems that have been successfully used by the Museum of Comparative Zoology for imaging insects - other imaging hardware and software could certainly be used, and may be more appropriate for other taxonomic groups.

1. Specimens >= 15 mm long

• Nikon D1X digital camera with NikkorMicro 105 mm lens
• Illumination by incandescent light (large specimens and those with high reflectivity)
• Illumination with the Microptics ML-1000 system (small specimens, those with low reflectivity, and microscope slides)

2. Specimens < 15 mm

• Nikon Coolpix 4500 digital camera mounted on a Leica MZ7.5 stereomicroscope


• Illumination with a Fostec fiberoptic light source

3. Specimens <15 mm that require full depth of field image

• JVC KY F75 digital camera mounted on a Leica MZ12.5 stereomicroscope
• Image data fed directly to Syncroscopy Auto-Montage image processing system


• Illumination with a Fostec fiberoptic light source

With the first two systems, captured images are virtually ready for incorporation into the image database, requiring only size modification and change of file format, both processes largely automated. For the third situation, each final image is the result of a digital assembly of a stepped series of images of the same structure taken at different focal distances, using the Auto-Montage image processing software.

Digital Documention Procedure

1. The type specimen is located in the collection and its label data are compared with the entry in the type database. Even with a complete database, occasionally type specimens may be absent from the registry, such as specimens recently returned or donated to the collection, or specimens not initially recognized as primary during compilation of the database. Also, species in the type database may be a syntype series, in which case the specimen in best condition should be selected for photography. In the event that another specimen is subsequently designated the lectotype by a specialist, that specimen should be photographed as well.


2. The specimen and all its labels are photographed - each imaged specimen receives a label indicating that it has already been documented.


3. Raw images, captured as high resolution, archival-quality TIFF files or low compression jpeg files, are assigned unique codes and placed in directories named with a combination of the species original name and its unique catalog number.


4. Large image raw files (TIFF and low compression JPEG files) are converted into smaller, lower resolution JPEG images for incorporation into the online database. Two JPEG versions are made: small (max. 500 pixels wide) and large (max. 900 pixels wide). Both the original raw files and processed smaller files are stored in two copies of hybrid format (readable on any platform) CD ROM's and a single copy data DVD. This ensures that there are at least three independent copies of each image, stored at different locations. A database of raw image files can be maintained in addition to the taxonomic database where processed images are stored. It may also be prudent to store image databases at remote sites, similar to the Harvard University Library's centrally managed Digital Repository Service. These services typically include back-up, monitoring to prevent format obsolescence, and strategies for data migration. It's entirely conceivable to have massive databases that exceed 1000 GB.


5. JPEG versions of images are linked to individual species records in the type database and assigned captions describing the structure or shot.

Obviously this endeavor will require a well-trained technical staff with competence in imaging and computers. The average number of type specimens a properly trained technician is capable of documenting in one day varies greatly depending on the taxon, size of specimens, number of required shots, the need to create composite images, and the type of equipment used. For example, photographing relatively large animals such as butterflies, fish, frogs and crabs requires no composite imaging, and one can expect to document between 30 to 50 species a day. At the other end of the spectrum, very small specimens, such as many ants, gastropods, flies and small crustaceans require multiple images (between 5 and 20) for each final composite photograph. In such cases only 4 to 14 species per day can be documented. Typically, 3 to 5 image views of each specimen are captured, in addition to associated labels. This includes dorsal, ventral, lateral and frontal views. Certain groups (e.g. some scarab beetles) require up to 8 individual specimen images if such action is suggested as a requirement for positive identification. Close-ups of the head, mouth parts, legs or other areas may then be needed.

Click here for more information on family-specific requirements.

top