Toward Transferable Water Level Trend Monitoring Using River Cameras: Integrating Domain-Specific Models and Segment Anything Model (SAM) for Water Segmentation

Water level variations influence the biochemical and hydrological processes within rivers. Through extensive river camera networks, obtaining reliable water segmentations from image data can practically support the monitoring of water levels. However, limited annotated data and tedious local deployment restrict the applicability of water segmentation models in new river scenarios. To pursue transferability, this study proposes a novel framework that combines domain-specific models with General AI for water segmentation. The framework utilizes a ResUnet model pretrained on a non-local dataset to identify pixels with the highest probability of being water. The Segment Anything Model (SAM), a promptable foundational computer vision model developed by Meta AI, is then employed to use these pixels as prompts for generating water masks. Different prompt modes of SAM are employed and compared. We applied the framework to image sequences acquired from river cameras stationed at four locations in Tewkesbury, UK. The framework significantly improved segmentation performance, with an increase of over 15% in Intersection over Union (IoU) over ResUnet. Meanwhile, the results substantiated point prompt as the more optimal mode for feeding prior knowledge to SAM. The static observer flooding index (SOFI) time series calculated based on the framework’s segmented masks under point prompt mode exhibit an average correlation of 0.90 with real water level fluctuations, significantly surpassing the single ResUnet model’s correlation of 0.54. Our study thus represents a step toward implementing river cameras for robust water level trend monitoring.