| Abstract Detail
Tools, Standards, Techniques, and Methods for Using Herbarium Specimens in Phenological Research Stucky, Brian [1]. Integrating and extracting phenology data from herbarium specimens with machine reasoning and learning. Herbarium specimens are a rich source of historical information about plant phenology, but extracting and digitizing this information at large scales is extremely labor-intensive. Furthermore, such data are often heterogeneous and difficult to integrate with other sources of information about plant phenology. In this talk, I present ongoing work to address both of these challenges. First, I will discuss recent efforts to develop a machine reasoning system and knowledge base, built around the Plant Phenology Ontology, that can “understand” different kinds of phenology data, enabling automated integration of disparate herbarium phenology data with phenology data from other sources. Second, I will discuss preliminary results of experimental efforts to use computer vision and machine learning techniques to automatically extract phenology data from images of herbarium specimen sheets. Log in to add this item to your schedule
1 - University of Florida, Florida Museum of Natural History, 1659 Museum Rd, Gainesville, FL, 32611, USA
Keywords: none specified
Presentation Type: Symposium Presentation Session: SY3, Tools, Standards, Techniques, and Methods for Using Herbarium Specimens in Phenological Research Location: 103/Mayo Civic Center Date: Tuesday, July 24th, 2018 Time: 8:45 AM Number: SY3003 Abstract ID:970 Candidate for Awards:None |