Automatic text location in images and video frames | IEEE Conference Publication | IEEE Xplore