Extracting Notes From Youtube Video

Abstract

In the world of technology, every B.tech student in our college uses Youtube to study the night before the exam, it will be of better use if there is an algorithm that can also provide notes for that video to look for last-minute revision. Hand-written notes are to be carried in a notebook, tedious to maintain for a long time. If a student wishes to pause and take notes, it will consume double the regular time of that streaming video. Those notes might or might not be used in the future or maybe not give proper context if the original content of the video is forgotten. In this project we are extracting frames from a presentation-based youtube video after every fixed time interval, using optical character recognition techniques on these images to make notes of the content written in a youtube video. Users will give the link of the desired youtube video and a pdf/doc file containing all the sentences shown in a video will be in the output. Students will just have to edit and delete some extra material that is not needed in notes. In the upcoming era education will bend more towards paperless, so to maintain the standards all the notes will be generated using this concept. Though it may sound simple it is a bit complex as to know when to extract data from given frame or how to handle cases when some person come in between the video. All this challenges will be major challenges in the future and will require a solution our idea is the first step towards this future.