INPAINTING SCHEME FOR TEXT IN VIDEO : A SURVEY

Mukesh Sakle, Rahul P. Chauhan

Abstract: Text data present in video sequences provide useful information of paramount requirement. Although text present in video provide useful information not all are of them are necessary because it may hide the important portion of the video. So there must way to erase this type of unwanted text. This can be done in two phase first text components are detected from each frame of the video. Detected text component are then removed from the video sequences. And restore the occluded part of the video using inpaint method. This text detection and removal scheme is in two phases. Each phase is broad topic of image processing. Video text detection and Inpainting are two most important phase in this scheme. Text detection phase consist of text localization, text segmentation and recognition phase. Inpainting method is used for restoring occluded part produce due to removal of text

Keywords: Optical Character Recognition(OCR), Stroke width transform, Text Detection,Connected Component

DOI: https://doi.org/10.15623/ijret.2014.0312032