A multiple frame integration and mathematical morphology based technique for video text extraction
Authors:
- 1 December 2010
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 434-437
Abstract
Text superimposed on the video frames provides supplement but important information about video content. Recognizing artificial text embedded in video frames is very important and meaning for video indexing and retrieval. In this paper, an efficient approach of text extraction based on temporal information and mathematical morphology is presented. In the first phase the blurred text is filtered, and then these clear text blocks are integrated in order to enhance the text. In the second phase, the edge map is got by using mathematical morphological edge detection. Lastly the missed text pixels in former phases are recalled using empirical rules. Experimental results show that our approach can effectively segment text from background and demonstrate the effectiveness of the proposed method in comparison with other related works.Keywords
This publication has 6 references indexed in Scilit:
- Accurate video text detection through classification of low and high contrast imagesPattern Recognition, 2010
- Using Multiple Frame Integration for the Text Recognition of Video2009
- Fast and robust text detection in images and video framesImage and Vision Computing, 2005
- Localizing and segmenting text in images and videosIEEE Transactions on Circuits and Systems for Video Technology, 2002
- Automatic caption localization in compressed videoIEEE Transactions on Pattern Analysis and Machine Intelligence, 2000
- AUTOMATIC TEXT LOCATION IN IMAGES AND VIDEO FRAMESPattern Recognition, 1998