Automatic Extraction of Thickness Information from Sub-Surface Acoustic Measurements of Manganese Crusts

Mapping and estimating the volumetric distribution of cobalt-rich manganese crusts (Mn-crust) is a challenging task that lies at the centre of deep-sea mineral prospecting. Acoustic methods are effective and capable of in-situ continuous measurements of Mn-crust thickness, providing much higher spatial resolutions compared to traditional methods involving sampling. However, processing acoustic signal in order to estimate thickness values is difficult due to low signal to noise ratios. This paper proposes a combination of image processing techniques in addition to acoustic signal processing in order to improve the accuracy of measurements. The advantage is the possibility of using the physical properties of Mn-crust, such as local continuity in order to recognize valid measurements. Testing the algorithm on data collected from sea experiments demonstrate that the reflected signals from the crust can be identified, resulting in spatially continuous thickness estimates.