Progress in brain research is widely based on experimental approaches having the potential to advance the knowledge on fundamental biology. In particular, the increasing availability of data obtained by experimental techniques has boosted the development of data-driven mathematical models reproducing neuronal circuits. While cellular and subcellular resolution data have been extensively collected in animal models, experimental constraints have limited the acquisition of information about human neuronal cytoarchitectures and morphologies. In this context, high resolution data can be efficiently gathered using non-linear optical techniques, such as two-photon microscopy, that, combined with fluorescent probes allows to describe with submicrometric resolution the complex morphologies and distribution of cells within intact tissues. Here, we present the results of a preliminary investigation performed on mouse brain slices to standardize experimental protocols that will be translated to acute human surgically excised tissues. More precisely, we have developed an immunofluorescence protocol for post-fixed brain slices of 300 microns to identify inhibitory and excitatory neuronal populations, using the NeuN monoclonal antibody and the GAD1 polyclonal antibody. Moreover, the experimental approach has been complemented with an image processing algorithm specifically dedicated to the segmentation of cell bodies that has been developed for the automatic identification of neuronal placement in 3-D. The proposed experimental and computational pipeline will be further validated against human samples providing an advanced and valuable tool to investigate the mesoscale organization of the human brain.