Skeleton analysis of population images for detection of isolated and overlapped nematode C.Elegans

In biotechnology, using image processing techniques to extract suitable information from C.Elegans nematode worms has recently drawn an increasing research interest. Many important problems in this area have not been solved yet, especially detection of each individual in population images. In this paper, we propose a method to detect individuals in population images. Our method is based on three steps: Pre-processing where a smooth skeleton is obtained from the image, Splitting skeleton into branches and angle calculation of each branch and finally merging the branches in order to detect each individual in the image. The outcome of the proposed method is a vision algorithm with 7.9% False Rejection Rate (FRR) and 8.4% False Acceptance Rate (FAR) on a database of 255 isolated and overlapped worms.