A method aligns a second image of an object taken at a second time with a first image taken at a first time, wherein the object experiences motion relative to an imager between the first and second times. The images are formed of sets of pixels arranged horizontally and vertically in a pattern. The method operates on the pixels of the second image by performing a series of operations to affect an affine transformation without performing matrix arithmetic, preferably by table look-up. The method shifts the pixels of the second image vertically by a constant vertical offset, shifts the pixels of the second image horizontally by a constant horizontal offset, shears the second image horizontally by a skew factor, shears the second image vertically by a skew factor, scales the second image horizontally by a horizontal scale factor; and scale the second image vertically by a vertical scale factor.