Low-Cost, Close-Range Digital Photogrammetry for Coastal Cliff Deformation and Beach Monitoring

  • Mitchell Eboigbe

Student thesis: Doctoral Thesis


Global warming, environmental change and sea-level rise are now perceived as a real threat to coastal communities across the World, including Wales, where coastal flooding and erosion have become more widespread and accepted. Existing techniques for monitoring the coastal environment are generally very expensive, in particular, remotely sensed datasets like LiDAR or satellite images. Consequently, many governments pay less attention to monitoring coastal areas and this has led to poor management programmes, resulting in the loss of lives and properties. In the United Kingdom, the Shoreline Management Plan does not necessarily encourage active intervention in protecting against coastal flooding and erosion. Photogrammetry is the most adopted technology for shoreline monitoring and management. While there are existing efforts in the application of close-range digital photogrammetry in monitoring Coastal Areas, there is no widespread adoption of a low-cost, consistent and regular monitoring methodology for data capture; processing and analysing that would accommodate the coastal cliff as well as accurately monitoring beach accretion and erosion. A cliff monitoring infrastructure is required to understand the cliff change pattern and correctly predict cliff falls. This research develops some new technologies for coastal cliff deformation and beach monitoring which are: A new pole-photogrammetry technique for regular cliff surveys; a new algorithm for coastal cliff deformation monitoring; an innovative predictive model for detecting and quantifying cliff fall and collapse; an innovative low-cost computational analysis and visualization for Beach Modelling and Monitoring. The new pole-photogrammetry technique is low-cost, accurate, rapid, and can generate high-resolution images. The methodology includes the most practicable survey distance to the cliff, vertical camera angle, and the best time Interval for executing a pole survey using the Integrated Sensor Orientation (ISO) from very cheap single frequency inbuilt GNSS. The new python algorithm for analysing the structural changes on the coastal cliff point cloud incorporates some existing libraries such as the Open3d, Numpy, and Pandas. This innovative algorithm reads every point in both point clouds and detects the changes. The algorithm creates three new point clouds for; where change has taken place, where there is no change, and the change array itself. The generated change-array point cloud was for computing the change volume and developing the coastal cliff prediction model. The new prediction model generates a probabilistic segmentation of all critical faults on the cliff and correctly predicts the date and quantity of cliff fall and collapse. The innovative low-cost computational analysis and visualization for Beach Modelling and Monitoring includes best practices; for low-cost and accurate UAV surveys, opensource and low-cost processing of digital images to acquire accurate orthomosaic and digital surface models, and for analysing digital images using open-source photogrammetric software. A three-stage monitoring procedure commencing with the small-scale photogrammetric datasets and ending in the critical evaluation of the dataset from the drone using the open-source QGIS was able to detect changes as small as 1cm to 3cm. The datasets from the drone and the pole integrate into a coastal monitoring infrastructure for monitoring the coastal area. The 1cm vertical accuracy and super-high-resolution images show an effective low-cost methodology for implementing strategic monitoring of the coastal cliff and beach. The recommendation includes a similar monitoring technique for the nearshore and coast and incorporates volume computation into the new change algorithm.
Date of Award2021
Original languageEnglish
Awarding Institution
  • University of South Wales
SupervisorDavid Kidner (Supervisor), Malcolm Thomas (Supervisor) & Nathan Thomas (Supervisor)

Cite this