The field of ghost imaging encompasses systems which can retrieve the spatial information of an object through correlated measurements of a projected light field, having spatial resolution, and the associated reected or trans-mitted light intensity measured by a photodetector. By employing a digital light projector in a computational ghost imaging system with multiple spectrally filtered photodetectors we obtain high-quality multi-wavelength reconstructions of real macroscopic objects. We compare different reconstruction algorithms and reveal the use of compressive sensing techniques for achieving sub-Nyquist performance. Furthermore, we demonstrate the use of this technology in non-visible and uorescence imaging applications. © 2013 SPIE.