Text extraction from cricket video- Comparative study of algorithms

Abstract

Video is one of the sources for presenting the valuable information. It contains sequence of video images & text information. Text data present in video contains useful information for automatic annotation, structuring, mining, indexing & retrieval of video. Nowadays manually added text in video sequences provides useful information about their contents. It provides supplemental but important information for video indexing & retrieval. To update the information on the website every one operator is required for watching and feeding data. To overcome these flaws we are using certain algorithms namely, Sobel algorithm, Prewitt’s algorithm, Robert’s algorithm & Canny edge detection algorithm. We are studying the comparison between these algorithms. This project contains firstly the processing of the video then we identified key frames from minimize the number of video frames. Then the key images are converted into gray images for the efficient text detection. Generally, the superimposed text displayed in bottom part of the images in the cricket video. So, we cropped the text image regions in the gray image which contains the text information. Then we are applying the above mentioned algorithms for text extraction from the cricket video.The algorithm that takes less time is the most efficient to work with. Hence, there is no need of any operator to feed the data on the website.
Keywords: Image Processing; text extraction; text recognition; text in video; video retrieval; video annotation; cricket video; video summarization; video text information and superimposed text.