Multi-View 3D Dynamic Vessel Reconstruction Using HydroCAM Array

-
Mixed Online/In-Person

Imo Abasiekong
Master’s Thesis Defense
Ocean Engineering: Ocean Mapping 

Tuesday, December 1, 2026, 9:30 a.m.
Chase 130

Abstract

Cameras mounted on waterway infrastructure, such as bridges, offer a fixed view of passing vessels. This thesis presents HydroCAM, an array of fourteen cameras that uses photogrammetry to reconstruct a vessel’s visible shape above water in 3D without onboard equipment. The workflow combines learning-based image segmentation, structure from motion, and multi-view stereo. Laboratory tests used three 3D-printed vessel models. The best whole-vessel reconstructions had median surface errors below 1 mm after alignment to the reference models. When a vessel is too long to fit in view, HydroCAM uses the fixed background and surveyed markers to reconstruct overlapping sections in a common refernce frame before merging them. For one model, this approach raised the F-score at a 5 mm tolerance from 25.7% to 35.3% using the same six captures. The final reconstructions from sections had length errors below 1% for both textured models. The nearly featureless third model did not yield a usable reconstruction from sections. HydroCAM also estimated passing clearance, the vertical gap between the vessel’s highest point and the structure above it. Using a laboratory stand-in for the water surface, the system overestimated clearance by about 6 to 24 mm, partly because it missed parts of the masts. Capture during continuous motion and performance on actual water remain unvalidated.

Bio

Imo Abasiekong graduated from the University of Uyo, Nigeria, where he earned his B.S. in Geoinformatics and Surveying. As an undergraduate, he first processed hydrographic survey data during a 2018 internship and co-authored a bathymetric study of Woji Creek, experiences that pointed him toward ocean mapping. After graduating, he spent four years working as a surveyor in Nigeria, progressing from graduate surveyor to party chief before moving into offshore marine geomatics and hydrographic surveying. In 2024, he joined CCOM/JHC as a graduate research assistant to continue his education in ocean mapping and completed Category A hydrographic training in 2025. He is pursuing an M.S. in Ocean Engineering: Ocean Mapping. His thesis research, advised by Dr. Semme Dijkstra, focuses on HydroCAM, a multi-camera photogrammetric system for georeferenced 3D reconstruction of surface vessels. He plans to continue developing optical remote sensing methods for 3D reconstruction in maritime applications.

Flyer for Imo Abasiekong's master's defense
Categories