National University of Singapore researchers have developed a method that automates the process of detecting text within video frames and extracting it from the background of each recording. The researchers note that their method for extracting text from video files would aid the effort to catalog large video databases and offer more effective search capability.
The researchers developed an automated text recognition method that first processes video frames using masks that enhance the contrast between text and background. Then they developed a process to combine the output of two known masks to enhance text pixels without generating image noise. From the contrast-enhanced image, their method then searches for characters of text using a Bayesian classifier, which employs probabilistic models to detect the edges of each text character.
Software detects how each character in an image relates to its neighbors to form lines of text by scanning its surroundings for nearby characters, growing the text box until the end of the line of text is found, and then eliminating false-positive results by checking that identified text boxes conform to certain geometric rules.
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found