Schematic representation of the developed automated workflow for the protein translocation analysis. Canaliculi are identified in the images by foreground-background detection, followed by cleaning from noise using morphological operations and thresholding. Foreground regions are then skeletonized and pruned. Further, fluorescence intensity profiles are extracted perpendicular to the selected skeleton fragments. Profiles undergo the selection by empirically identified criteria and ranking according to the quality parameters. Zones are identified in average intensity profiles and numerical descriptors are evaluated. Wilcoxon rank sum test is applied to compare datasets.