Abstract: Extracting geometric primitives is an important task in the reverse engineering. We propose a new method for extracting cylinders from an unorganized set of 3D points. The originality of this approach is to separate the extraction problem into two distinct steps based on the speciality of the cylindrical Gaussian image. The first step consists in extracting the direction of the cylinders in the Gaussian image using a genetic algorithm. This yields a subset of 3D points along with a direction. In the second step, cylinders of known direction are extracted in the corresponding subset of points. Experimental results show that the cylinder scheme is effective and robust.