This article attempts to measure the effect of various noises present in scanned documents on the performance of various APIs in the OCR segment.