ALBERT

Description: Currently, the multitude of information about marine geophysical, biological data and spatial patterns increases tremendously. Compared to the increasing amount of information only a few concepts and techniques are applied for efficient visualisation and optimal analysis of present and upcoming data sets. There is a considerable need for ageneralised analysis of acoustic data, still photographs, video flows and video mosaics of the sea floor. For example, more than 4200 geo-referenced video mosaics have been derived during the RV Polarstern equipped with the remotely operated vehicle (ROV) Victor6000 cruise to the Håkon Mosby Mud Volcano (HMMV) in 2003. Therefore the automatic or semiautomatic image analysis gains more and more in importance concerning greyscale, colour, texture and contour detection.An automatic extraction of the areas of interest could be realized by edge detection andgeometrical shape recognition of the Freeman Chaincode algorithm. The transmission of the shapes into polygons allows a further spatial analysis within GIS and with geostatistical methods. This includes spatial budgets of geological and biogeochemical cycles and characterisation of habitats at the seafloor based on the combination ofseveral information layers.The geo-statistical methodology is used to be practiced to analyse structures and to describe spatial patterns. The classical tool, the interpolation, takes place in describing the variography of spatial dependence of data (semi-variance) as a function oftheir distance. This spatial variability is examined in order to derive a model which describes the spatial characteristics of the variables. This model is the crucial input for the actual interpolation, the Kriging. The Indicator Kriging used in this study operates with data from a bivariate data type to predict the probability of thresholds (e.g. occurrence yes (1) or no (0)) of the same data type at unsampled locations. Thus the kriged surfaces reveal whether the probability of occurrence of a certain habitat indicator e.g. bacteria mats 〈 50% is particularly high or particularly low. This can be accomplished successively for all interesting regions. Finally a potential habitat map of the benthic communities at HMMV results from the interpolation. Some surfaces can remain undefined, since the Kriging permits an interpolation only up to the distancedetermined by the semivariogram. Beyond that no statements about surfaces aremade. For validation of the computed results, Kriging allows calculation of standard errors of prediction. Furthermore, we applied a cross-validation analysis which givesinformation about the accuracy of the computed maps.The combination of semiautomatic video mosaic analysis and geo-statistical methods enables us to estimate the percentage of coverage by diverse parameters. The different information layers were overlaid by GIS technique for the consideration of different geochemical habitats, mud flows and calculation of spatial budgets.

Description: The combination of new underwater technology as Remotely Operating Vehicles (ROVs), high resolution video imagery, and software to compute georeferenced mosaics of the seafloorprovides new opportunities for marine geological or biological studies and applications in offshore industry. Even during single surveys by ROVs or towed systems large amounts of images are compiled. While these underwater techniques are now well-engineered, there is still a lack of methods for the automatic analysis of the acquired image data. During ROV dives more than 4200 georeferenced video mosaics were compiled for the Håkon Mosby MudVolcano (HMMV). Mud volcanoes as HMMV are considered as significant source locations for methane characterised by unique chemoautotrophic communities as Beggiatoa mats. For the detection and quantification of the spatial distribution of Beggiatoa mats an automated image analysis technique was developed, which applies watershed transformation and relaxationbasedlabelling of pre-segemented regions. Comparison of the data derived by visual inspection of 2840 video images with the automated image analysis revealed similarities, a precisionbetter than 90%. We consider this as a step towards a time efficient and accurate analysis of seafloor images for computation of geochemical budgets and identification of habitats at theseafloor.