ReefScan-Shallow is a flexible system for mapping and monitoring shallow water systems. The system was used to undertake semi-automated diver-less 2.5D image surveys of coral reef reef-flat environments. Source images were collected using a specialised camera boom housing six cameras, mounted on a tender that was self-driven in a lawn-mower pattern using an electric trolling motor controlled by a GPS. Each survey was 0.5Ha in size (typically 100m x 50m), from reef edge through to reef flat, capturing image data at millimetre resolution. The resulting images were geo-coded and stitched into large-scale ortho-mosaics using Agisoft Metashape™.
Twenty sites along the Great Barrier Reef were surveyed from 2021 to 2023. The data allows the spatial organisation of coral colonies over large areas to be measured and changes to individual coral colonies quantified. The ortho-mosaic can be analysed using Machine Learning models to identify and give percent cover of major benthic forms.
The system delivers hectare-scale capability to map reefs at a resolution that provides information about fine and larger scale patterns of corals, and changes to these, that sits between smaller-scale diver plots and larger-scale satellite data. Future work will develop the system into a fully automated platform with associated workflows.