The introduction to the paper is reproduced here:
"The use of embedded image metadata is becoming widespread in the biodiversity informatics community (e.g. Stafford et al. 2010 & Tulig et al. 2012), and is frequently used to describe the subject and licencing of images as well as for recording the 'tombstone metadata' (e.g. Introduction to Metadata) - when the image was created, last edited, who created it, and where and how it was created.
The eMonocot project (http://about.e-monocot.org) makes use of the Scratchpads (Smith et al. 2011) infrastructure as a tool for collecting, curating, and creating content to be harvested by the eMonocot portal (http://e-monocot.org). As part of this project hundreds of images with embedded metadata are being uploaded to a number of different Scratchpads, combined with images directly uploaded by partner communities, and exported en mass to the portal. For this to be technically feasible at scale images from varied, disparate sources need to have their metadata standardised as part of the bulk upload process.
There are three widespread image metadata formats that can be handled by this module. A subset of the EXIF standard (Camera and Imaging Products Association Standardization Committee 2010) specifies a method for tagging of images with metadata. This is widely used by device manufacturers to record both the make and model of the image capture device and also the device's settings when the image was captured (e.g. focal length, flash duration). The eXtensible Metadata Platform (XMP) was originally developed by Adobe Systems Incorporated and later adopted by the International Standards Organisation as ISO 16684-1:2012. It uses a data model defined in Adobe 2012 which is serialised in XML when embedded into files. The International Press Telecommunications Council defines the IPTC Core and Extension metadata standards (IPTC 2010).
EXIF Custom on drupal.orgAn existing Drupal module, Exif (https://drupal.org/project/exif), provides a mechanism for displaying embedded image metadata on Drupal nodes, but does not provide a mechanism for mapping the metadata into fields. The import of embedded metadata into Scratchpads/Drupal fields is a requirement of the eMonocot project and is useful for the wider Scratchpads community as it allows for these data to be easily used by other Drupal modules (e.g. Views - https://drupal.org/project/views) and in other Scratchpads-specific functions such as our on-going work on implementing the ability to export data via DarwinCore Archives (GBIF DarwinCore Archives). There is a comparison of these two modules (and potentially other similar Drupal modules) at https://drupal.org/node/1842686."
Read the full paper (Biodiversity Data Journal)