Automated system for text detection in individual video images

摘要: Text detection in video images is a challenging research problem because of the poor spatial resolution and complex back- ground, which may contain variety colors. An automated system for text presented. It makes use four modules to implement series processes extract regions from images. The first module, called multistage pulse code modulation (MPCM) used locate potential color converts image coded image, with each pixel encoded by priority ranging 7 down 0 accordance its priority, further produces binary thresholded segments background. second region applies sequence filters remove noisy eliminate that are unlikely text. third box finding merges boxes likely Finally, fourth optical character recognition (OCR) eliminates produce no OCR output. extensive set experiments conducted demonstrates proposed effective detecting wide © 2003 SPIE IS&T. (DOI: 10.1117/1.1584050)