Abstract: Nearshore underwater artificial structures (e.g., tunnels, pipelines) are susceptible to instability and damage due to long-term hydrodynamic impacts and construction activities, posing risks to operational safety. The decline in performance of such structures (exemplified by seawater intake systems), often caused by scouring corrosion, biofouling, or blockages, necessitates regular and precise inspection and assessment. Grounded in the concept of multi-source data fusion, this study proposes an acoustic-optical multimodal joint diagnostic methodology. By integrating three underwater monitoring technologies—multibeam bathymetry, dual-frequency imaging sonar, and optical imaging from remotely operated vehicles (ROVs)—an acoustic-optical-topographic comprehensive data evaluation framework is established. Innovative spatiotemporal registration and dynamic quantitative analysis methods for multi-sensor data are introduced. Field tests demonstrate that this approach enables accurate defect identification under complex marine conditions, significantly improves detection efficiency compared to conventional methods, and effectively mitigates reconstruction errors induced by scouring. The findings provide valuable insights for the intelligent operation and maintenance of marine engineering structures such as offshore wind turbine foundations and cross-sea tunnels.