Artificial Intelligence (AI) Driven Coastal Change Detection of Littoral Waters
Loading...
Authors
Orescanin, Mara
Young, Walt
Herrmann, Dave
Subjects
Advisors
Date of Issue
2018-04-18
Date
4/18/2018
Publisher
Language
Abstract
Unmanned aerial vehicles (UAVs) are cost-effective platforms that can be used to quantify large-scale coastal change caused by extreme events, such as beach breaching, typically caused by significant storm events. The U.S. Navy and Marine Corps often conduct exercises and real world operations in littoral waters and the accompanying coastal landscape. Damage to DoD assets resulting from the passage of such events is difficult to predict and therefore developing quick and accurate methods to assess change, both to the coastal landscape and to coastal infrastructure, is essential. This project uses digital aerial imagery (visual and IR) from UAVs in conjunction with in-situ measurements of water/sand properties to develop a techniques to monitor and quantify coastal morphological and water quality response with the passage of extreme weather events. Specifically, UAV surveys were conducted at Carmel River State Beach, Carmel, CA, that has undergone significant morphological change due to mechanical and natural river breaching events. The Carmel River is an ephemeral river, characterized by periods of beach closure during dry months and periods of direct connection between the river and coastal ocean during wet months. During the seasonal transition from dry to wet, the River undergoes a series of breaching and closure events. These events are unpredictable, and are similar morphologically to breaches caused by storms on barrier beaches. Given that the River and Beach morphology are constantly changing, this location provides the opportunity to test UAV monitoring techniques. Large area images of the beach are compiled from UAV surveys to provide digital elevation maps (DEMs) of the beach. These DEMs are currently being analyzed to determine the amount of sediment transport during the breach-closure cycle. In addition, a deep learning computer neural net is being developed and trained on a coastal dataset with the intention of creating a change detection algorithm that will detect areas of greatest change from pre-storm to post-storm morphologies.
Type
Presentation
Description
CRUSER TechCon 2018 Research at NPS. Wednesday 3: Applications
Series/Report No
Department
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.