Identifying mining sites in conservation areas from satellite imagery
Mentor: Dr. Fei Fang and Dr. Afsaneh Doryab
Client: Dr. Fei Fang, Carnegie Mellon University
Mining in conservation areas can be environmentally damaging, and is a serious challenge in countries such as DC Congo. Miners' presence in the conservation area leads to additional threats to wildlife and an increase in illegal logging activities. Given the vast area in need of protection, it is often difficult for the law enforcement agencies to get up-to-date information about the mining sites in the conservation areas. Fortunately, satellite imagery can serve as an important data source that can provide timely information about such threat.
In this project, the students will work on designing algorithms to automatically identify mining sites from satellite imagery through leveraging advanced machine learning techniques. In addition, with the identified mining sites over the past few years, the students will analyze their dynamics, such as how the locations of the new mining sites changed over time. The designed algorithm can be used to monitor new mining sites in conservation areas, which will assist the conservation agencies to identify new threats to wildlife and forest.