Robust extraction of text in video | IEEE Conference Publication | IEEE Xplore