Bangor University: 3D scan of a saltmarsh

The terrestrial laser scanner, or TLS, represents the latest tool in remote sensing. With dimensions similar to that of a large suitcase, the TLS uses precise laser light to return a panoramic 3D point model of the surrounding landscape. The end scan is detailed (50,000 points per second), far ranging (with a range of 300 meters) and rapid (scans take less than 5 minutes).

Cai Ladd of Bangor University and Tim Durban-Jackson of SEACAMS (associated with the School of Ocean Sciences) have worked on a laser scan survey of a local marsh in North Wales. The survey took about 10 hours and was done using a Leica Scanstation C10 by placing the scanner at multiple positions in the marsh. The video shows the amazing scope of this technique for examining marsh features across multiple scales.

TLS has been used in several topographic applications such as stratigraphic modelling (Bellian et al., 2005), geomorphic change detection (Schürch et al. 2011) and habitat mapping (Watt et al., 2005).

Though impressive, there are limitations to the technique. Points are only generated for an object if it is in direct line of sight of the laser, and light return is lower further away from the scanner. An end model will therefore have a lot of “shadow” and a sparser point cloud at its periphery.

A collaboration between SOS and SEACAMS is attempting to combine TLS output with another survey technique called “Surface through Motion”, to compensate for this shortfall. The technique uses parallax effect in a series of overlapping images to generate 3D models from 2D photographs. In combination with TLS, this hybridised approach is hoped to generate accurate and “complete” 3D models of a given landscape.

Furthermore, the collaboration hopes to use vegetation height, texture and spectral signatures obtained from the scans as metrics for habitat classification. If successful, this could offer a novel tool for resource managers to rapidly conduct vegetation surveys of a habitat.

Cai Ladd, Bangor University PhD student (read Cai’s PhD summary and blog)