A system for processing a video signal into a processed video signal includes a pattern recognition module for detecting a region of text in the image sequence based on coding feedback data and generating pattern recognition data in response thereto. A video codec generates the processed video signal and generates the coding feedback data in conjunction with the processing of the image sequence.