In the reconstruction of facade elements, the identification of specific object types remains challenging and is often circumvented by rectangularity assumptions or the use of bounding boxes. We propose a new approach for the reconstruction of 3D facade details. We combine MLS point clouds and a pre-defined 3D model library using a BoW concept, which we augment by incorporating semi-global features. We conduct experiments on the models superimposed with random noise and on the TUM-FACADE dataset. Our method demonstrates promising results, improving the conventional BoW approach. It holds the potential to be utilized for more realistic facade reconstruction without rectangularity assumptions, which can be used in applications such as testing automated driving functions or estimating facade solar potential.