In recent years, in the context of the application of Multivariate Curve Resolution (MCR) to hyperspectral image analysis, attention has been more and more put onto the possibility of exploiting not only the spectral but also the spatial information for constraining the algorithmic solution. Examples involve the introduction of different spatial constraints during the iterative Alternating Least Squares (ALS) calculation of the MCR solution or the post-processing of the score images using conventional image processing techniques. In this framework, this work proposes an approach for constraining concentration distribution maps within MCR-ALS analysis of hyperspectral images, based on the use of spatio-spectral fuzzy clustering in order to obtain smoother, more contrasted, and better interpretable chemical images. We show the relevance of the proposed approach and investigate the effect of the application of a spectral-spatial fuzzy clustering constraint on samples of different nature.