27 Mar 2014

Automated Pollen Counting

From Our Changing World, 9:20 pm on 27 March 2014

Green kiwifruit pollen, and kat Holt with a Classifynder

Kat Holt with a Classifynder that is identifying two different pollen grains on a slide: green kiwifruit (left) and lycopodium (right) (images: A. Ballance)

Palynology – the study of pollen - which has until now has been a very time-consuming research method, is on the brink of a revolution. Kat Holt, from Massey University’s Institute of Agriculture and Environment, shows Alison Ballance the Classifyinder, a smart automated system capable of learning to identify and count different kinds of pollen. The Classifynder has been developed by the Palynology Automation Research Group, based in the School of Engineering and Advanced Technology at Massey University, led by Emeritus Professor Bob Hodgson and pollen expert Emeritus Professor John Flenley.

Kat and researchers in a handful of research organisations around the world are currently teaching Classifynder how to identify different pollen types, and developing library files of views of different pollen grains.

You can listen here to an interview with John Flenley about his pollen research on Easter Island.