The microscopic Raman imaging method was used to observe the working surface of a grinding wheel used in the grinding of silicon wafers. Raman images of the working surface of the grindstone were generated using peak feature values of the Raman peaks—peak intensity, peak position, and peak—with the 1330 cm^-1 and 520 cm^-1 bands as labels for object identification. The results showed that each object could be identified from their Raman images under the conditions used for analysis of the working surfaces consisting of mixtures of diamond abrasives, resin binders, and generated residues, such as single-crystal silicon chips, modified amorphous silicon materials, and grindstone wear debris.