- Development of an Integrated UAV-based 3D Mapping and Change Detection System for Coastal Erosion Monitoring

Project Abstract

Coastal erosion poses significant threats to infrastructure, ecosystems, and local communities, necessitating advanced monitoring systems that can deliver timely, accurate, and actionable geospatial information. This research proposes an integrated UAV-based 3D mapping and change detection system to monitor and quantify coastal erosion dynamics with high spatial and temporal fidelity. The system combines multi-sensor data fusion, including high-resolution RGB, multispectral, and LiDAR-like depth information obtained via structure-from-motion and dense stereo techniques, to generate centimeter- to decimeter-scale 3D models of dynamic coastal zones. A robust workflow was developed to address challenges such as rapid shoreline movement, variability in surface textures, tidal effects, and atmospheric conditions. The methodology includes flight path optimization and calibration procedures to maximize data overlap, a photogrammetric pipeline for dense point cloud generation, and a LiDAR-inspired height normalization framework to ensure consistent vertical measurements across campaigns. Change detection is achieved through a multi-temporal comparison of fixed shoreline features and volumetric analyses of dune, beach, and cliff morphologies, employing both raster-based and vector-based approaches, as well as machine learning classifiers to distinguish natural erosion from anthropogenic alterations. The integrated platform supports near-real-time processing by leveraging parallel computing, cloud-based storage, and an interactive visualization interface that enables stakeholders to review, validate, and interpret results efficiently. Accuracy assessments conducted against terrestrial LiDAR surveys and ground control points demonstrate improved shoreline delineation accuracy and volumetric change estimation, with mean vertical errors under 5 cm for critical dune features and horizontal deviations within 10 cm under favorable conditions. The system was tested across multiple coastal settings with varying geomorphologies, including sandy beaches, barrier islands, and cliffed coasts, to evaluate generalizability and robustness to environmental noise. Results indicate that the proposed framework can detect shoreline retreat, dune toe migration, and cliff recession with high confidence, enabling early warning and evidence-based decision-making for coastal management, disaster risk reduction, and habitat preservation. Sensitivity analyses reveal the most influential factors driving measurement uncertainty, such as sea state, cloud cover, and sensor calibration drift, informing recommended practices for repeat surveys and long-term monitoring programs. The research also explores integration with existing geographic information systems (GIS) and coastal management dashboards, facilitating data interoperability, scenario analysis, and stakeholder engagement. Practical implications include improved allocation of protective measures, optimized dune restoration planning, and enhanced documentation of erosion processes for regulatory compliance. Limitations identified pertain to extreme weather conditions, rapid sediment redistribution events, and access constraints that may affect survey cadence, suggesting avenues for future work such as automated flight scheduling, adaptive sampling strategies, and incorporation of hyperspectral materials analysis to better classify sediment types and moisture content. Overall, the study contributes a comprehensive, scalable, and interpretable solution for continuous coastal erosion monitoring using UAV-based 3D mapping and change detection, with potential application to climate adaptation planning and coastal resilience initiatives.

Project Overview

What This Project Is About

---

A plain-language overview of using drones (UAVs) to create detailed 3D maps of coastlines and detect changes over time, such as erosion or sediment movement. It blends simple mapping, data comparison, and clear results to understand how coastlines evolve and where protection may be needed.

---

---

The Problem It Addresses

---

Coastlines are constantly changing due to tides, storms, and human activity. Traditional surveys are time-consuming and may miss small, important changes. This project develops an approachable way to capture accurate 3D coastal models and automatically spot differences over time to support planning and conservation.

---

---

Objectives of the Project

---

1. Create a workflow to capture high-quality drone imagery of a coastal area.
2. Build a 3D model of the coast from the images.
3. Develop a change-detection method to identify shoreline and cliff changes between surveys.
4. Validate results with simple ground-truth checks and existing data.
5. Provide an easy-to-read report and visualizations for stakeholders.

---

---

What You Will Do Step by Step

---

1. Learn basic drone operation and safety requirements for coastal work.
2. Plan survey sites, collect aerial images, and record metadata like weather and time.
3. Process images to build 3D coastal models (textures and elevations).
4. Run change-detection analyses to compare models over time.
5. Interpret results, create simple maps and charts, and draft a final report.

---

---

Expected Outcome

A practical, user-friendly workflow that delivers 3D coastline models, clear change maps, and actionable insights for coastal management and conservation.