Construction and validation of economic vision system for bacterial colony count

This study aimed to construct and validate the vision system for bacterial colony count on agar plates. The constructed vision system comprises of hardware components and software. The hardware components include a 4,000 lux light control box and a digital camera. The software was designed for user friendly operation, therefore; the interesting region can be selected; and threshold and colony diameter can be adjusted. Staphylococcus aureus, Escherichia coli and Bacillus cereus, important foodborne pathogens which possess different colony morphology were used for colony counting by the constructed vision system. The validation results revealed that the average percentage error was 2.13% in comparison to the counting by the expert. The constructed vision system has high precision and accuracy with a statistical significant. The Pearson correlation was 0.998 implying that the developed vision system has high performance for bacterial colony count. Thus this vision system is practically and economically for bacterial colony count in microbiological laboratories.