Based on the x-ray absorption edges of different elements, we simultaneously image and distinguish the composition of three differently shaped components of an object by using energy-resolved x-ray absorption ghost imaging (GI). The initial x-ray beam is spatially modulated by a series of Hadamard matrix masks, and the object is composed of three pieces of Mo, Ag, and Sn foil in the shape of a triangle, square, and circle, respectively. The transmitted x-ray intensity is measured by an energy-resolved single-pixel detector with a spectral resolution better than 0.8 keV. Through correlation of the transmission spectra with the corresponding Hadamard patterns, the spectral image of the sample is reconstructed, with a spatial resolution of 108 µm. Our experiment demonstrates a practical application of spectral ghost imaging, which has important potential for the noninvasive analysis of material composition and distribution in biology, medical science, and many other fields.