Over the past century, monitoring of Giardia lamblia became a matter of concern for all drinking water suppliers
worldwide. Indeed, this parasitic flagellated protozoan is responsible for giardiasis, a widespread diarrhoeal disease (200
million symptomatic individuals) that can lead immunocompromised individuals to death. The major difficulty raised by
Giardia lamblia's cyst, its vegetative transmission form, is its ability to survive for long periods in harsh environments,
including the chlorine concentrations and treatment duration used traditionally in water disinfection. Currently, there is a
need for a reliable, inexpensive, and easy-to-use sensor for the identification and quantification of cysts in the incoming
For this purpose, we investigated the use of a digital holographic microscope working with partially coherent spatial
illumination that reduces the coherent noise. Digital holography allows one to numerically investigate a volume by
refocusing the different plane of depth of a hologram.
In this paper, we perform an automated 3D analysis that computes the complex amplitude of each hologram, detects all
the particles present in the whole volume given by one hologram and refocuses them if there are out of focus using a
refocusing criterion based on the integrated complex amplitude modulus and we obtain the (x,y,z) coordinates of each
particle. Then the segmentation of the particles is processed and a set of morphological and textures features
characteristic to Giardia lamblia cysts is computed in order to classify each particles in the right classes.