Self-dual Attribute Profiles for the Analysis of Remote Sensing Images

The spatial relations are essential information that should be considered when analyzing remote sensing images. Attribute profiles (combinations of an anti-granulometry and a granulometry computed with connected operators based on attributes) can be employed for the modeling of the spatial information of the surveyed scene. In this paper we propose self-dual attribute profiles which are attribute profiles computed on an inclusion tree with self-dual operators. The proposed variant of the attribute profile was effectively considered for the classification of a very high geometrical resolution remote sensing image.