Hierarchical feature representation of geospatial objects using morphological pyramid exploitation

This paper presents a novel hierarchical feature representation for geospatial objects detection from optical very high resolution (VHR) satellite imagery. An information extraction approach using multi-scale and hierarchical exploitation for remote sensing image is presented. The key idea is that we adopt a morphological pyramid-based framework for geospatial objects detection in VHR imagery, with a combination of morphological pyramid exploitation and a novel hierarchical feature representation metric, in order to develop an efficient procedure for multi-scale analysis of geospatial object detection to address the issue of information uncertainty in practical applications. We test our method on optical VHR QuickBird satellite imagery and obtain promising experimental results, which confirm the effectiveness and robustness of the proposed procedure.