The demand for efficient stone detection techniques of stony areas (reefs) in the marine environment increased in the last years. This is especially important for seafloor areas with a patchy distribution of stones in an otherwise sand-dominated milieu. These hard substrates are hotspots of marine biodiversity; especially for benthic communities. For fishes, marine mammals and birds stony areas are important breeding and feeding places. Current research projects are dealing with the stone distribution and density to investigate e.g. species-habitat interactions. On the political level detailed distribution maps are relevant for resource assessments, coastal management and protection conventions. The detection of stony habitats in sublittoral environments is still a considerable challenge in spite of modern high resolution hydroacoustic techniques. Object detection on the seafloor is commonly done on the basis of hydroacoustic backscatter intensities recorded with e.g. sidescan sonar and multibeam echo sounder. Single objects such as stones can generally be recognized by the acoustic shadow behind the object. For small areas single objects can be easily identified manually; for large scale mapping automated techniques are required which are still under development. We collected a set of hydroacoustic data on the German continental shelf to compare and discuss the following approaches to detect and demarcate stony habitats: • manual detection on backscatter mosaics • automated detection on backscatter mosaics using machine learning techniques • seismo-acoustic approach (sediment echo sounder, sidescan sonar)