ADVANCED TEXT CLASSIFIER BASED ON THE CONCEPT OF PIPELINING AND CC CLUSTERING

Image processing is a way to alter an image into digital structure and carry out a few operations on it, in order to obtain an improved image or to take out a little valuable information from it. In this paper, testing has been made to pull out text from remote sensing images. The task is to separate text from the non-text area of an image. The separated text is converted into speech. And the way of segregation of text and translating into speech is supportive in lots of ways, such as guiding visually impaired persons; help to navigate tourists, etc., the citation of text area from a remote sensing image is done by maximally stable extremal region algorithm and the Ada boost classifier. The overall process of text extraction has to be done and achieved through the major concept called pipelining. Varied algorithms have been used in implementing this technique. Traditionally Region-based algorithm was used and it had lots of hindrance mainly the decisions were made locally. Hence connected component has been introduced which reduces the disadvantages of Regional-based algorithm to a certain extent. The main purpose of the experiment is to overcome the disadvantages like photometric degradation, geometric distraction. The result is to extract text area with accuracy and speed and also to overcome the disadvantages of previous techniques.