During my talk I will introduce our (2 + 1)D vicinal Cellular Automaton model and show that it is a powerful tool in the analysis of different structures on the crystal surface. I will show that the diffusion of particles in the inhomogeneous potential energy landscape leads to many unexpected and interesting behaviors of the system. The combination of the direct and inverse step barrier and the proper selection of the well potential between them or the change in the height of the step barrier leads to the growth of nanocolumns, nanowires, and nanopyramids or meanders in the same system. I will show that not only the combination of step barriers is crucial in their formation. The surface characteristic we also investigate through analysis of the two length scales: height and horizontal distance. I will present that proper identification of length scaling allows for pattern identification in a simple, automatic way.