Automatic Colony Counter using Computer Vision and Artificial Intelligence

Bacterial colony counting is a critical step in biomedical research and clinical diagnosis. The manual counting process usually carried out by health analysts will only count a quarter of the number of colonies in the petri dish. Due to the large number of bacterial colonies that appear on the petri dish. However, this has a big effect on the final result of the analysis. The goal of this research is to create a real-time, automated method for counting bacterial colonies based on artificial intelligence and computer vision. YOLOv5 TensorRT on Jetson Nano, is used in this research to create and test deep learning models for automated microbial colony detection. This research demonstrated the results of computations performed manually and those performed using colony counter tools are compared, the average value is 0.0142%. This is considered an excellent number since the lower the error value obtained, the more exact the instrument.