object recognition methods as preprocess of object counting in BAL medical images

pattern recognition has been radically developed theoretically and practically during the last decade. Nowadays effects of this appfoach are remarkable in many fielsd like health care industrial etc. one of the most important problems inmedicine is object boundaries detection and variety of objects in medical images there are so many methods for solving this kind of problems the applicability and effectiveness of each approach depends on the image quality and tissue type this paper proposes three pattern recognition methods together with their results analysis and their comparison . after all practical fields of each method are discussed. proposed methods briefly are: color based method FCM1 based method and histogram based method.

کلیدواژه ها:

FCM ، histogram ، medical image ، object recognition ، object counting