Under the influence of global climate change, many large-scale disasters, such as floods, typhoons and other large-scale disasters, so that in recent years to study the issue of disaster prevention, has gradually been concerned about. For this technology is still in the initial stage, but Taiwan is rare for this technology. In this case, this paper uses remote sensing and spectroscopy, and detect and quantify areas where high-risk areas and potentially disasters occur in Taiwan. Through different algorithms cross-comparison and statistics, do a step closer analysis. However, the use of these new technologies, develop a system for detecting bare land in this areas. Using this system to predict disaster. In order to reduce the damage caused by the disaster.