Automatic Delineation of Burned Forest Areas from Satellite Imagery to Analyze and Manage Wildfires

Delineating burned forest areas is an important first step to analyze predict and prevent wildfires. Numerous methods to delineate burned areas were published about this topic already. However, none has focused on its automatic applicability, yet. For many research projects, such as our recent project KIWA, analyzing forest structures and their exposure to fire risk is important. Thus, identifying and delineating burned areas to later-on identify typical landscape and forest structure patterns within them is vital to understand these structures. The higher the degree of automation together with a high reliability and simple applicability, the easier fireprone areas can be identified and predicted. Against this background we were investigating some meaningful approaches for an automated and easy to handle method to delineate burned areas.