Optical flow guided cell segmentation and tracking in developing tissue

Cell segmentation and tracking is necessary for analyzing cell motion in time-lapse data. In this paper, we present a method which combines optical flow tracking and temporal consistency modeling to segment and track cells in time-lapse data. The method is designed for tracking cells in a dense (developing) tissue, e.g., kidney tubule in Xenopus, where cell candidates can be obtained from frame-wise segmentation methods. By modeling the temporal consistency, we can select the most probable configuration as the result of segmentation and tracking.