Gram-staining characterisation of activated sludge filamentous bacteria by automated colour analysis.

An automated image analysis method has been developed for the monitoring of the Gram-staining characteristics of filamentous bacteria in activated sludge. The binary method of pixel classification agreed with manual estimation (level of correlation of 0.9 for Gram-positive bacteria). Its robustness has been assessed by repeatability tests. Population shifts in terms of Gram-staining characteristics have been monitored in laboratory-scale experiments with two feeding schedules using this technique.