IIIF BGBM Example
Part of IIIF Pages
Since the beginning of digitisation at the BGBM in 1999, we have relied on zoomable image formats and software. The image format fpx was used for this. In the course of time, we had to use different commercial server software systems for on-the-fly provision of images, as there has been a discontinuation of software products over the last 20 years. Furthermore, there was hardly any software capable of both running in the cloud and managing extensive image inventories locally. One of the main goals here was to enable variable image display in the different resolutions needed without having to do any prior image processing.
Three years ago, we started to provide all BGBM specimen images (~400,000) via an iiif server in parallel to the existing systems. The migration to iiif-only image delivery was completed in early 2020.
Implementation
Currently we use the IIIF server digilib for IIIF display of our images. The use of digilib gave us the possibility to use our existing image formats (tif, jpg) directly without having to do extensive conversions, because digilib can use jpeg directly via a plugin. The diglib server is integrated into our internal image management system as a publication server.
In our portals we use the Mirador 2 viewer and the OpenSeadragon viewer. At the moment, all our IIIF outputs are implemented in APIs version 2.1. The creation of the manifest files is done in the context of content negotiation when requesting the corresponding endpoint. The underlying services and functions have been developed using the SLIM PHP framework based on scripts developed at the RBGE Edinburgh (Code available in GitLab).
We have recently started publishing the manifest file via our BioCASe Provider Software installations (see section Linking with ABCD Metadata) and we are testing the presentation in external portals such as GBIF and Europeana. The individual fields in the manifest file correspond to dwc fields and are filled on-the-fly from JACQ. We are offering this functionality to other herbaria that also manage their specimen data in JACQ as a service for making their associated images available via IIIF.
Future development
As next steps we want to make further image sources (autographs, living specimens, archivals) available via IIIF services. We also plan to implement access to our IIIF elements via searches by name or person similar to the capabilities developed in Edinburgh.
The implementation of version 3 of the IIIF API is also under way.
Back to IIIF Home
