Marine hard substrates are a rare habitat type in many shallow coastal shelf areas, such as the North Sea. They form the vital fundament for a multiplicity of sessile organisms and, further, provide shelter, nursery and feeding ground for mobile species, inter alia for ecologically and economically important fish species. However, despite the importance of hard-substrate habitats, they face numerous anthropogenic pressures. Economic activities such as fishing, sediment extraction, exploitation of oil and gas, and construction of off-shore wind farms can have a direct influence on the habitat structure and the functioning of the ecosystem. Besides this, the effects of climate change, pollution and eutrophication, sedimentation, and the introduction of invasive species entail adaptation or loss and lead to changes of the epibenthic assemblages. Within European waters hard substrates are protected by the European Habitats Directive (habitat type reef, code 1170), nevertheless, there is still the requirement to map and monitor these habitats in order to implement strategies towards a sustainable use of marine resources and an effective protection of sensitive areas. Obtaining robust and reproducible base date for these purposes is paramount and conditional upon the use of adequate methods. The aim of this thesis project was to provide an approach towards automatic detection of individual hard substrates on high-resolution spatial side-scan sonar (SSS) mosaics, which is suitable for the demarcation of reefs. It was also aimed to characterize hard-substrate habitats in the German Bight (south-eastern North Sea) with respect to the distribution and colonization of individual stones based on underwater video footage, and to determine and to describe the epibenthic assemblages in consideration of biotic and abiotic processes. They main results show, that the distribution of hard substrates varies within the German Bight and that the dislocation of mobile sediments seems to have a strong influence on the presence of hard substrates and the structure of associated epibenthic assemblages. This is especially relevant for cobbles located in the nearshore areas. These sediment-mediated disturbances lead to a dominance of long-lived sessile taxa on boulder-sized stones, whereas short-lived taxa can be readily found on cobbles. It was shown that Haar-like feature algorithms can be used to detect stones on SSS mosaics, as long as a sufficient amount of training samples are available. In summary, the data obtained and analyses carried out for this thesis provide new insights into the spatial distribution, characteristics and functioning of an important marine habitat type in the south-eastern North Sea. The data illustrate that hard substrate habitats are presently neither fully documented in their diversity and heterogeneity (in space and time), nor sufficiently well understood in terms of a functional ecosystem. These data are highly relevant for applying and constructing monitoring strategies and modelling of this habitat type, facilitating the maintenance of optimal habitat conditions in a changing environment.