Endothelial junctions are dynamic structures organized by multi-protein complexes that control monolayer integrity, homeostasis, inflammation, cell migration and angiogenesis. Newly developed methods for both the genetic manipulation of endothelium and microscopy permit time-lapse recordings of fluorescent proteins over long periods of time. Quantitative data analyses require automated methods. We developed a software package, the CellBorderTracker, allowing quantitative analysis of fluorescent-tagged cell junction protein dynamics in time-lapse sequences. The CellBorderTracker consists of the CellBorderExtractor that segments cells and identifies cell boundaries and mapping tools for data extraction. The tool is illustrated by analyzing fluorescent-tagged VE-cadherin the backbone of adherence junctions in endothelium. VE-cadherin displays high dynamics that is forced by junction-associated intermittent lamellipodia (JAIL) that are actin driven and WASP/ARP2/3 complex controlled. The manual segmentation and the automatic one agree to 90 %, a value that indicates high reliability. Based on segmentations, different maps were generated allowing more detailed data extraction. This includes the quantification of protein distribution pattern, the generation of regions of interest, junction displacements, cell shape changes, migration velocities and the visualization of junction dynamics over many hours. Furthermore, we demonstrate an advanced kymograph, the J-kymograph that steadily follows irregular cell junction dynamics in time-lapse sequences for individual junctions at the subcellular level. By using the CellBorderTracker, we demonstrate that VE-cadherin dynamics is quickly arrested upon thrombin stimulation, a phenomenon that was largely due to transient inhibition of JAIL and display a very heterogeneous subcellular and divers VE-cadherin dynamics during intercellular gap formation and resealing.