Quantitative comparison of two particle tracking methods in fluorescence microscopy images

Abstract

Tracking of multiple bright particles (spots) in fluorescence microscopy image sequences is seen as a crucial step in understanding complex information in the cell. However, fluorescence microscopy generates high a volume of noisy image data that cannot be analysed efficiently by means of manual analysis. In this study we compare the performance of two computer-based tracking methods for tracking of bright particles in fluorescence microscopy image sequences. The methods under comparison are, Interacting Multiple Model filter and Feature Point Tracking. The performance of the methods is validated using synthetic but realistic image sequences and real images. The results from experiments show that the Interacting Multiple Model filter performed best, under the test conditions.