A method of forming a composite image using a CMOS image sensor. The method comprises capturing a plurality of frames using the image sensor, identifying a reference point in each of the frames, and aligning the frames using the reference point. Finally, the frames are combines, such as be an arithmetic addition, into the composite image.