An Approach for Subtitles Extraction from Videos and Converting to Text Document

With the increment and wide use of practical vision systems and smart phones, text detection in videos has become a critical yet challenging task. In today’s world and advancement in easily available and different featured cameras, text recognition has become a tough work and changes are occurring in this field. Texted content present in the video provides useful information likewise scene understanding, for narration, it adds details to the music playing or video playing. With the expansion of digital libraries therefore there is intense need of a method to retrieving the image using the embedded content in them. The proposed system is to detect text and extract them from the videos containing subtitles. The algorithm proposed here aims to get better subtitle extraction form these texted frames of videos. Our approach is to extract the subtitles from the videos by using edge detection and clustering and stroke width transform.