The very first map of the Arctic Ocean basin with a few lead line sounding changed the supposal of large continental land beneath of the ice. More resolution added over the decades, reveals the detail of the Arctic seafloor structure of seamounts and ridges below the frozen sea. Numerous methods of bathymetry and mapping were applied as the technology developed over the years for different purposes. While the airborne and satellite-based altimetry and gravimetry data provides a large-scale estimation of the seafloor topography by hundreds of meters resolution, the shipborne and submarine sonars focuses on certain features and areas with higher resolution. During the last century the knowledge of the Arctic seabed geomorphology increased dramatically by the development of acoustic technology combined with altimetry and gravimetry while the habitat characteristic of the polar region still contains lots of mysteries. The new developments of underwater survey vehicles are bringing new clarity and perspective from the deep sea to the questioners. The sub-meter resolution data of the seabed could be employed for very high-resolution micro topography as well as habitat mapping and feature detection. The Alfred Wegner Institute for Polar and Marine Research (AWI) developed the Ocean Floor Observation and Bathymetry System (OFOBS) for deep sea research, mostly in polar region. The tailored deep tow system of the AWI is equipped with optical and acoustic sensors in addition to underwater positioning systems. The OFOBS, first deployed during the PS101 expedition, provides a novel dataset of megafauna’s habitats at the Karasik seamount. This thesis is implementing geospatial data mining and knowledge discovery for feature detection by means of habitat mapping in the study area with a focus on the central mount of Karasik seamount where an imperial assemblage of the Geodia sponges are dominating the seafloor. The main datasets for this study are based on the optical sensor of the OFOBS, including video and still images collected during the dives, while the feature detection within the sonar dataset is in the second place. During this work study, the development of the OFOBS is also considered in order to improve the capability of the dataset for further expeditions.