This paper presents a method for vectorizing the graphical parts of paper-based line drawings. Th e method consists in separating the input binary image into layers of homogeneous thickness, skeletonizing e ach layer, segmenting the skeleton by a method based on random sampling, and simplifying the result. The seg mentation method is robust with a best bound of 50% noise reached for indefinitely long primitives. Accurate estimation of the recognized vector's parameters is enabled by explicitely computing their feasibility doma ins. Theoretical performance analysis and expression of the complexity of the segmentation method are derive d. Experimental results and comparisons with other vectorization systems are also provided.