Introduction to Extraction of Text from Video:

In these modern days everything is online like shopping, education, jobs, music, games, news and many more. To connect with the people far from us through video chat or video calling and many us are also e learners that is they learn by accessing internet and also music lovers watch video songs. Everyday life is circulating around internet only from the wake up to the sleep. In accessing internet we may come across plenty of videos. Sometimes we may not able to understand the voice from a video or we want to extract the text from the video for better understand.

For example consider a Traffic control police uses a digital camera and catches the video of somebody who crossed traffic limits and he wants to see the vehicle number of that person. Here comes the problem of extracting texts from the videos. Not only a traffic police or a police come across these problems, a common people may also come across these kinds of many problems in their day to day life.

This project can solve the above problem of extracting text from a video using video text detection and recognition. By using edge detection based method we can extract the text from video and use matlab for implementation. Here we use the concept of critical angles 90 degree called threshold angles and 180 degree and the text is extracted and by using the edge image concept we recognize the text. Extracting the text from a video is a two step process of finding text region of the original image and separation of text from the background.

This project finds many real time applications like semantic video indexing, video surveillance, summarization and in many more. But the main drawback of this application is that it works only on the images having proper edge.

Download  Extraction of Text from Video ECE Seminar Report.