The stability of W/O/W double emulsions is governed by the structure of the interfacial stabilizer: polymeric stabilizers such as poly(vinyl alcohol) (PVA) form adsorbed layers with pronounced steric hindrance and interfacial viscoelasticity, whereas monomeric surfactants such as Tween 80 adsorb rapidly but form comparatively weaker, less elastic interfacial films. These differing interfacial architectures are expected to govern osmotically driven water transport and droplet-scale morphological stability, yet quantifying this relationship at scale has been limited by conventional image analysis, which struggles with concentric inner/outer boundaries and requires manual annotation. This study presents an automated pipeline based on the Segment Anything Model 2 (SAM2), a zero-shot segmentation foundation model, applied to microfluidic time-lapse video of W/O/W droplets to quantitatively compare PVA- and Tween 80-stabilized systems. Frames are CLAHE-preprocessed before automatic mask generation, and candidate masks are filtered by solidity, aspect ratio, and area. To distinguish genuine concentric inner/outer pairs from duplicate detections of a single boundary, we developed a deduplication algorithm combining centroid proximity, relative size, ellipse IoU, and a ring-brightness gate that separates dark duplicate inner-boundary detections from bright genuine inner/outer pairs. The pipeline automatically classifies each droplet as Inner or Outer and extracts per-frame position, radius, and aspect ratio without manual intervention. Preliminary application to PVA-stabilized video confirms high-throughput, reproducible tracking with distinguishable Inner/Outer populations. Ongoing work applies this pipeline under matched conditions to PVA- and Tween 80-stabilized systems to test whether steric and viscoelastic differences between polymeric and monomeric stabilizers yield measurable differences in droplet size stability and inner-phase dynamics; results will be presented.