not without image classification and change detection being done. You could do this on a pixel level, but this is of little use to a human... it is the aggregation of pixels into groups of 'changed here' that may ultimately arise from machine learning