作者: Martin Maska , Arrate Munoz-Barrutia , Carlos Ortiz-de-Solórzano , None
DOI: 10.1109/ISBI.2012.6235805
关键词:
摘要: We present a fast and robust approach to tracking whole fluorescent cells in time-lapse series. The proposed scheme involves two steps. First, coherence-enhancing diffusion filtering is applied on each frame reduce the amount of noise enhance flow-like structures. Second, enhanced cell boundaries are detected by minimizing Chan-Vese model level set-like framework. To allow simultaneous multiple over time, contour evolution has been integrated with topological prior exploiting object indication function. Preliminary experiments 2D series GFP-transfected adipose-derived stem demonstrate high accuracy short execution times.