Title: Text Detection Tracking and Recognition in image using Connected Components
Authors: D.Kavya, A.Mohan Krishna
Volume: 2
Issue: 8
Pages: 140-143
Publication Date: 2018/08/28
Abstract:
The study of image text is presently in wide demand because an image is a main source of afferent data in our lives. Text is a prominent and direct source of information in an image. But is a great challenge for us in order to detect, extract and recognize the text by seeing an image due to the difference in size, style, orientation, robustness to background complexity, text degradation and distortion, text variations, and moving objects. There are major contributions in order to detect, track, and recognize the text in an image. First one is a generic framework for text extraction. Second includes of methods and systems are summarized, compared and examined and final one is related applications, prominent challenges and future directions for text extraction. In order to extract the text, we have an algorithm based on two machine learning techniques(classifiers). One to make candidate word regions, other eliminates non-text regions. These educed CCs are divided into small parts known as clusters so that we can produce candidate regions. In additional to MSER algorithm we have AdaBoost classifier that discover the closeness relational and cluster CC’s by their pair wise relations. Defines whether each region contains text or not