Automatic segmentation and counting of Skeletonema costatum obtained by flow cytometry image

Abstract

Skeletonema costatum (S. costatum) is one of the most common species that can give rise to red tide, a global marine disaster. For the purpose to monitor and forecast the outburst of red tide caused by S. costatum, many measures have been brought up. In this paper we propose a recursive method to segment and calculate S. costatum images captured by our improved Flow Cytometry. Firstly Otsu method is applied to separate the algae cells from the background. Afterward, the flood filling algorithm is adopted to fill the holes produced by the Otsu method. Then we use the erosion procedure to recursively find out the contours of each S. costatum cells. The result shows that our method can correctly separate most of the cells.