Abstract: In computer vision the extraction of text in an image is a classical problem . Extraction process includes detection, localization, tracking, extraction, enhancement and recognition of the text from the given image. The problem of automatic text extraction extremely challenging because of variation of text due to difference in size, style, orientation, alignment, low image contrast and complex background make. Text extraction needs binarization which leads to loss of important information contained in gray scale images. Extraction becomes more difficult because the images may contain noise and have complex structure. This paper involves an algorithm which is insensitive to noise, skew and text orientation. It does not contain artifacts that are generally introduced by thresholding using morphological operators. There are many examples that presents the performance of proposed method.

Keywords: Mathematical Morphology, Localization, Morphological Operators, Connected Component, Edge Detection