Traditional hard classification of remote sensing images has been proved to be inaccurate due to the presence of mixed pixels in images.Even though the composition of these pixels for different classes can be estimated with pixel unmixing model the output provides no indication of how such classes are distributed spatially within these pixels.Sub-pixel mapping is a technique designed to obtain the spatial distribution of these classes in these pixels with information contained in mixed pixels.A newly Cellular Automata model was proposed to solve the problem of sub-pixel mapping with the assumption of spatial dependence.The model was tested on both synthetic and real images

and the result shows that this Cellular Automata model is a simple and efficient method to solve the sub-pixel mapping problem.