This paper proposes a new approach for automatic extraction of roads from SAR imagery based on genetic algorithm. We first suppress the speckle by Frost filter

then detect line feature pixels to get potential road pixels. Followed that

extract primitive line segments by Radon transform on every set of connected pixels

and use genetic algorithm to group line segments to get the candidate road segments. Finally

The snake model is used to adjust the position of the candidate road segments in order to make road segments discrimination using the road’s linear feature more effective. We show the feasibility of the approach not only by presenting reasonable results on airborne and spaceborne SAR images but also by evaluating them quantitatively based on ground truth.