Journals Proceedings

Abstract

Digital video now plays an important role in entertainment, education, and other multimedia applications. Artificial text in video is normally generated in order to complement or encapsulate the visual content and thus is an important transporter of information that is highly significant to the content of the video. Problem arises when the content of video getting more in time and user facing the struggle to search and edit the desired information. The user cannot get or copy the artificial text directly from video but need to retype the text content in the video. This will consumes a lot of times and work. The main objective of this project is to extract the artificial text in Chinese Simplified and English appear in video and save it into word format. This paper proposed an effective and efficient caption extraction from video. Proposed methods which comprised of video segmentation, image de-nosing, image segmentation and optical character recognition (OCR) are used to extract artificial text in the video. The recognition percentage of text character in this project is up to 97.8%. The recognized text is saved in word format. The video containing different language can be extracted to obtain the text for future work. This system can be improved by extracting the scene text in the video.