The extraction of characters from cover images using mathematical morphology

Abstract

In understanding an image, the extraction of charcters existing in the image is considered to be important. Scene images are differ from document images, which are composed of characters and complicated background (i.e.photo, picture, or painting etc.) instead of white one, this make it difficult to be dealt with. In this paper, we introduce a new method to extract characters from scene images using mathematical morphology. In this paper, a new approach to extract characters from scene image is proposed. For a scene image, there are many complicated information on it. The information carried through characters is considered to be most important among them. therefore, the extraction of characters from scene image in order to recognize them become increasingly concerned.[l]-[3] The characters on the scene images are difficult to be extracted, because are intricate with complicated background, and show various size, shape, direction and situation. Although the importance of this field have led a number of researchers to pay attentions to it and there have been several studies in the character extraction, are proposed to be optimal ideas for extraction characters, but are limited to simpler background or other restrictive conditions. In this paper, a new method using mathematical morphology is proposed. Mathematical morphology provides us the theory and the tools to analyze shapes, while the decomposition of shapes plays an important role in the implementation of morphological operation. This effect has been noticed by many researchers in the shape based processing and several optimal algorithms [4]-[8] have been proposed. Characters in scene images may be characterized by word, they float on the background or sink un