We develop and apply comprehensive and robust methods to extract and classify natural objects from continuous and discrete raster datasets. Relevant features are acquired to describe changes in landscape and land resources at different levels using image data. Mathematical-statistical methods are adopted for automatic detection and description of image objects. Thus we contribute concepts, methods and data to describe/detect area wide changes and processes in the resources of landscape.

Tasks and main research:

• Development and application of methods to extract natural objects from continuous data.

• Development of methods for a comprehensive description of natural and anthropogenetic boundaries in continuous pattern (e.g. map signatures, vegetation transition, forest borders).

• Development and application of methods to extract 3D-information from remotely sensed data for description of natural structures and changes. The main focus lies on wood and its embedding/interaction within/with the landscape.

• Conception and development of data acquisition based on high resolution remote sensing data.

• Conception, development and maintenance of the software interface in area wide data acquisition using airborne remote sensing data.

• Scientific expert advice and support in the fields of photogrammetry and survey at WSL. Maintenance, enhancements and future development in these specific fields.

• Adequate presentation of scientific results on national level and in noted international journals and at international congresses/workshops/symposia.