Semi-automatic PFU Counting Method by Double Thresholding and Two-Stage Segmentation

Abstract

Virus quantification is the necessary step for measuring the vaccine efficiency. Plaque assay is the most popular method and quantified the virus in term of a Plaque Forming Unit (PFU). In this paper, a semi-automatic method for plaque assay from image taken by a mobile phone is proposed. Two thresholds in different color spaces are used to extract the PFUs: one for PFU at the center and one for PFU at the edge of a well. The cluster of PFU is segmented in two stages. In the first stage, watershed segmentation based on distance transform is applied to segment the distinct and large PFUs. In the second stage, the size is used to separate the blurry PFUs. The proposed method achieved 94% precision, 92% sensitivity and 6% counting error in the experiment on 30 samples stained with neutral red color.