This document summarizes an algorithm for blind source separation using independent component analysis (ICA). ICA is used to separate mixed images into their original independent components without knowing the mixing process. The proposed algorithm first preprocesses the data through centering and whitening. It then uses an iterative approach to maximize the non-Gaussianity of the independent components, extracting them one by one through deflation. Simulation results on mixtures of 5 images show the algorithm can effectively separate the images, with peak signal-to-noise ratios for the recovered images similar to the originals.