CYWG iDigBio Image Ingestion Appliance Use Cases
Use-case scenarios for iDigBio image ingestion appliance
Concurrently with the specimen portal, the iDigBio IT team is developing a software appliance to facilitate the ingestion of digital images into the iDigBio cloud by authorized users. Ingestion of an individual image is a core primitive that the iDigBio cloud’s server-side Application Programming Interface (API) exposes to users; the image ingestion appliance is primarily concerned with the client-side software that builds upon the API primitive to enable reliable transfers of batches of images; it also exposes a simple Web-based interactive interface. The use-case scenarios that have driven its design include:
1) Allow end-users to ingest images to the iDigBio cloud to provide a mechanism to support crowd-sourcing
One example of this usage scenario is a multi-institution TCN where images are digitized at multiple sites, later to be crowd-sourced (e.g., to extract specimen data from digitized labels).
The appliance in this scenario provides the ability for users distributed across the TCN to independently:
- ingest batches of images,
- assign GUIDs to media information,
- make the image searchable on the portal through media information, and
- generate lists of media information for each batch including the HTTP accessible endpoint at iDigBio for each ingested media object.
This list can then be shared with other users affiliated with the TCN, who can then easily access images online through their browser, shared with a crowd-sourcing application that can make use of the images with a broad base of citizen scientists, and/or used to load the data into the TCN data management system to enable the linking of the image with specimen data.
2) Allow end users to ingest images that are linked to specimen records
One example of this usage scenario would be an institution that already has Audobon Core (media information) associated with media files. By uploading the media files and media information (including the user-provided GUIDs for the media, and the user-provided GUID for the specimen), these images can be searchable by media information or through the linked specimen information.
3) Provide a basis for integration with third-party bio-collections tools to create appliances that can automatically ingest images into the iDigBio cloud
One example of this usage scenario would be to enhance an existing bio-collections tool that manages databases and image repositories for a collection, so that both specimen data and images can be together ingested/updated in iDigBio, though a push-based approach rather than pull-based. Integration of image ingestion code with a bio-collections tool is necessarily tool and programming language dependent, but two general strategies are envisioned:
- to create libraries or services that can be accessed through tool “plug-ins”, or
- to integrate the bio-collections tool with the image ingestion code as a single virtual appliance as in use case 2, where the bio-collections tool and the image ingestion code are coupled through the generation of intermediary files (using standard terms and file formats).