Rugosity, slope and aspect from bathymetric stereo image reconstructions

This paper demonstrates how multi-scale measures of rugosity, slope and aspect can be derived from fine-scale bathymetric reconstructions created using geo-referenced stereo imagery collected by an Autonomous Underwater Vehicle (AUV). We briefly describe the 3D triangular meshes generated from the stereo images and then present a detailed overview of how rugosity can be derived by considering the area of triangles within a window and their projection onto the plane of best fit. By obtaining the plane of best fit, slope and aspect can be calculated with very little extra effort. The results are validated on a simulated surface and the effects of mesh resolution and window size are explored. The technique is demonstrated on real data gathered by an AUV on surveys that cover several linear kilometres and consist of thousands of images. The ability to distinguish habitat types based on rugosity and slope are demonstrated through K-means cluster analysis. A human labelled data set is then used to train a SVM classifier that exhibits promising habitat classification potential based on rugosity and slope.