Road Extraction from SAR Multi-Aspect Data Supported by a Statistical Context-Based Fusion

In this paper we describe a fusion approach for automatic object extractionfrom multi-aspect SAR images. The fusion is carried out by meansof the Bayesian probability theory. The first step consists of aline extraction in each image, followed by attribute extraction.Based on these attributes the uncertainty of each line segment isestimated, followed by an iterative fusion of these uncertaintiessupported by context information and sensor geometry. On the basisof a resulting uncertainty vector each line obtains an estimationof the probability that the line really belongs to a road.     «

In this paper we describe a fusion approach for automatic object extractionfrom multi-aspect SAR images. The fusion is carried out by meansof the Bayesian probability theory. The first step consists of aline extraction in each image, followed by attribute extraction.Based on these attributes the uncertainty of each line segment isestimated, followed by an iterative fusion of these uncertaintiessupported by context information and sensor geometry. On the basisof a resulting uncertainty vector eac...     »