An Algorithm for Individual Intermediate Filament Tracking

摘要

In this paper we propose an algorithm for automatic tracking of individual intermediate filaments (IF), which form a highly branched network inside living cells. The algorithm is based on Stretching Open Active Contours (SOAC) combined with a novel method for processing the snake's open ends. The main element of our approach (Endpoint-Controlled Active Contours) is a newly introduced potential function, which is included into the snake's energy term forcing its end-points to coincide with the filament's end- or junction-points. The comparison of our method with similar tracking algorithms for the problem of IF tracking shows a significantly improved accuracy of the overall tracking.