Detection of stationary vehicles in airborne decimeter resolution SAR intensity images using morphological attribute filters

The increasing resolution of modern SAR sensors leads to new approaches for object detection in SAR images. Objects appear not just as blobs but rather show a certain structure which can be exploited by advanced approaches already established in image processing. But an increase in detail comes at the cost of an increased complexity. In this work we use Mathematical Morphology (MM) to detect stationary vehicles in airborne decimeter resolution SAR intensity images. Mathematical Morphology has become popular because of its proven utility and sound mathematical foundations. Here we use the concept of connected components to simplify the image and to describe parts of the vehicles. The detection is a subsequent process. It is shown that a robust detection based on the explicit description of the vehicle appearance is possible.     «

The increasing resolution of modern SAR sensors leads to new approaches for object detection in SAR images. Objects appear not just as blobs but rather show a certain structure which can be exploited by advanced approaches already established in image processing. But an increase in detail comes at the cost of an increased complexity. In this work we use Mathematical Morphology (MM) to detect stationary vehicles in airborne decimeter resolution SAR intensity images. Mathematical Morphology has be...     »