Automated Analysis of Planktic Foraminifers, Part II: Image Acquisition
Abstract
Manual collection of a statistically significant quantity of unbiased and reproducible data on microfossil morphology is time consuming. Therefore, fully automated operating robots are needed that recognise and analyse microfossils. Here, a robot is described that autonomously measures the size and shape of randomly oriented planktic foraminifers using an incident light microscope. The robot captures about 1000 images of particles larger than 63 micrometers per hour and then extracts prominent morphological features including size. The system has successfully completed about 800 deep-sea sediment analyses with >1.6 million size measurements of planktic foraminifers (see Part I, Schmidt et al., this volume). The robot can also be used for automated analysis of particles other than planktic foraminifers because elementary problems such as contrast and focus on an incident light microscope have been overcome using a low reflectance and levelled glass tray. The system is currently developped for automated recognition of planktic foraminifers using a neural network classifier (see Part III, Schiebel et al., this volume).
- Publication:
-
EGS - AGU - EUG Joint Assembly
- Pub Date:
- April 2003
- Bibcode:
- 2003EAEJA....12450B