Baryon Acoustic Oscillations (BAO) are a feature imprinted in the density field by acoustic waves travelling in the plasma of the early universe. Their fixed scale can be used as a standard ruler to study the geometry of the universe. BAO have been previously detected using correlation functions and power spectra of the galaxy distribution. In this work, we present a new method for the detection of the real-space structures associated with this feature. These baryon acoustic structures are spherical shells with a relatively small density contrast, surrounding high density central regions. We design a specific wavelet adapted to the search for shells, and exploit the physics of the process by making use of two different mass tracers, introducing a specific statistic to detect the BAO features. We show the effect of the BAO signal in this new statistic when applied to the Lambda - Cold Dark Matter (LCDM) model, using an analytical approximation to the transfer function. We confirm the reliability and stability of our method by using cosmological N-body simulations from the MareNostrum Institut de Ci`encies de lEspai (MICE). We apply our method to the detection of BAO in a galaxy sample drawn from the Sloan Digital Sky Survey (SDSS). We use the `Main catalogue to trace the shells, and the Luminous Red Galaxies (LRG) as tracers of the high density central regions. Using this new method, we detect, with a high significance, that the LRGs in our sample are preferentially located close to the centres of shell-like structures in the density field, with characteristics similar to those expected from BAOs. We show that stacking selected shells, we can find their characteristic density profile. We have delineated a new feature of the cosmic web, the BAO shells. As these are real spatial structures, the BAO phenomenon can be studied in detail by examining those shells.