Table 1.

Web services to enable the implementation of workflows for processing digital herbarium specimens (23, 24)

Category	Subcategory	Name	Input	Output	Description
Image-based	Object recognition	Scale Matching Service	GUID, SRI	Scale region coordinates	Template matching algorithm (uses an example image of the searched for scale to detect it in other images) (23)
Image-based	DPI recognition	DPI Service	SRI resolution, SRI size, scale region coordinates	DPI	Computation of DPI using the physical size of the actual scale and the size of its digital counterpart on the herbarium sheet
Image-based	Object recognition	Text Region Service	GUID, DPI	Text region coordinates	Line contrast approach (taking advantage of the fact that the horizontal contrast of text lines is very high and dark and bright areas are alternating quickly)/machine learning approach to detect text-like structures
Image-based	OCR	Tesseract/OmniPage Service	GUID, text region coordinates	Text	Tesseract/OmniPage OCR algorithm
Text-based	Dictionary-based	Scientific Name Extractor	Text	Scientific names	Parsing with dictionary based on Global Names (http://gnrd.globalnames.org/) and WikiData (24)
Text-based	Dictionary-based	Botanist Name Extractor	Text	Botanist names	Parsing with dictionary based on botanists’ database of the Harvard University Herbaria & Libraries (http://kiki.huh.harvard.edu/databases)
Text-based	Regular expression	Date Extractor	Text	Dates	Matching using regular expressions for dates related to collection, accession and determination
Text-based	Regular expression	Geographical Coordinates (GeoCoord) Extractor	Text	Coordinates	Matching using regular expressions for geographical coordinates
Text-based	Dictionary-based	Location Extractor	Text	Locations	Parsing based on Cartographic Location and Vicinity Indexer (CLAVIN) library (http://clavin.bericotechnologies.com)

GUID indicates Globally Unique Identifier; SRI, Scale Reference Image.