Iron ore pellets are formed by an agglomeration process and currently constitute the main source for the reduction process in steel making. The fractions of solid phases and pores directly affect pellets´ properties such as compression resistance, gas permeability during the reduction process, and reducibility. In this work a method for the automatic identification and quantification of phases and pores in iron ore pellets was developed, based on the correlation between images obtained with two different techniques – optical microscopy (OM) and scanning electron microscopy (SEM). Mosaic images covering the full equatorial cross section of a pellet were acquired with OM and SEM. Employing digital image processing techniques the phases and pores were identified and quantified in each type of image. However, each imaging technique has limitations in the discrimination of certain phases, preventing a full quantification. On the other hand, the combination of the two types of images allows discriminating all phases. For that, OM and SEM images were automatically registered using homologous reference points obtained with the SIFT – Scale Invariant Feature Transform technique. After registration, phases and pores were individually identified and quantified, leading to much more accurate results than those provided separately by OM or SEM. The porosity was also compared with that provided by x-ray MicroCT. For that, a correlation procedure identified the closest matching MicroCT layer to the OM or SEM images, the image was registered and the pore fraction was measured. The obtained value is much lower for the MicroCT image, what was attributed to the worse spatial resolution of the technique.