T201 The Future of Quantitative Paleontology: Biometry, Computer Vision, and Machine Learning
Tuesday, 6 November 2012
1:30 PM - 5:30 PM
217A
QUANTIFYING THE COMPLEXITY OF BIOLOGICAL SHAPE
Paper number: T201-4
Program: Technical Sessions
Time: 2:15 PM – 2:30 PM
Authors
-
Mao LiFlorida State University, Tallahassee, FL
-
Luke ManderUniversity of Illinois at Urbana-Champaign, Urbana, IL
-
Surangi W. PunyasenaUniversity of Illinois, Urbana, IL
Abstract
We develop computational methods for quantifying the complexity of shapes that exhibit a high level of morphological variability and apply the techniques to the analysis of the patterning of the granules on the exterior wall of pollen grains. We employ these techniques in conjunction with methods of image processing and pattern analysis to develop an algorithmic approach to the taxonomic classification of several species of grasses based on morphological and textural characteristics observed in high-resolution images of pollen grains acquired through scanning electron microscopy.