Machine vision application to the detection of water-borne micro-organisms

Abstract

The work presented in this paper uses a novel Machine Vision application to detect and identify micro-organism oocysts on drinking water. This new concept of water borne micro-organism detection uses image processing to allow detailed inspection of parasite morphology to nanometre dimensions. The detection results are more reliable than existing manual methods. The machine vision proposed provides a 100% detection of cryptosporidium micro-organism as test case. Combining Normarski Differential Interface Contrast (DIC) and fluorescence microscopy using Fluorescein Isothiocyanate (FITC) and UV filters, the system provides a reliable detection of micro-organisms with a considerable reduction in time, cost and subjectivity over the current labour intensive time consuming manual method.