Each lexicon is compared with the image to find the best match between the radicals and the given string by substring alignment. In this process, the segmentation and recognition results are ...