Segmenting and tracking multiple deformable objects is a topic of popular interest in 3D biological imaging. In addition to quantitative measurements, a great importance lies in the visual observation of the results, usually available via time-consuming reconstruction methods, which introduce approximation errors. In this paper, we propose a framework based on deformable mesh models able to segment and track multiple objects simultaneously while providing precise 3D visualization of the model during its evolution. This active mesh framework is evaluated on simulated data, and experimental results on real biological images are shown.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below