Mining frequent substructures from deforestation objects

The fight against deforestation is a priority for the environmental organizations, society and government. It demands the creation of methodologies and techniques that allow monitoring and intervening, efficiently and at reasonable costs, in areas susceptible to deforestation. Thus, the computational modeling of remote sensing data involves many challenges, including a large set of algorithms and techniques to extract strategic information contained in these data. This paper aims to employ graphs to represent relationships among deforestation objects captured from remote sensing data, and then extract patterns from them applying graph mining, that performs the search for frequent substructures using the FSG graph-based knowledge discovery algorithm, in order to identify frequent substructures among deforestation objects.