An original local-potential morphometric algorithm for information-entropy estimation of the impact of technogenic chemical pollution on forests was proposed. This algorithm is distinguished by a procedure of separation of binary objects—pixel clusters (simply connected, tightly packed sets of white and black pixels)—in satellite photographs by a potential transformation and a procedure of estimation of the changes in the local configuration of pixel clusters by calculating the information entropy for the virtual potential of models of the interaction of pseudoparticles (white and black pixels) of the binary image. This enables one to practically estimate the spatial disturbances in forests by the impact of technogenic chemical pollution using point statistical estimates of the geometric indices of clusters. The proposed algorithm differs from conventional ones that determine the boundaries of binary objects by statistical analysis using co-occurrence matrices or point statistical estimates of the configuration of pixels.