A novel method for flash character localization

It has been a challenge to locate character in flash due to the variations of character in size, color and style. Besides, multilingual text in flash brings more difficulties to character localization. In this paper, a novel character localization method based on connected component analysis is proposed for locating character in flash. The proposed method first clusters the color to separate the color image into homogeneous color layers. Then a heuristic connected component merging algorithm is performed to merge some connected components which satisfy the characteristic of character in each color layer on the premise of non-merging mistakes. Next, for effective character localization, the features of characters are extracted by connected component clustering, and then most characters will be located according to these features. Finally, a new noise removal method is performed to remove all non-characters connected components, and then all characters are located. The experimental results show that the proposed method can locate flash character effectively.