Fuzzy Rank Morphology Theory And Its Applications In Image Processing

Mathematical morphology has been widely used in the field of digital image processing because of its advantage of flexibility, nonlinearity and because it's an ideal method for parallel processing. But most of the methods for extending the existing binary mathematical morphology theory to gray image field cannot maintain the dual property of it. In this thesis, based on the domestic and overseas researching, we have proposed the fuzzy rank morphology theory.First, we have studied about the connection between rank order theory and mathematical morphology, used the fuzzy logical theory to define fuzzy rank morphological erode and dilate operators which can keep the dual property of mathematical morphology.Second, based on the new morphological operators, we have defined the opening and closing operators and filtered contaminated images. Since that in binary mathematical morphology theory, combination of openings and closings can filter an contaminated image. Experimental results show that the new method for filtering a contaminated image performance better than the classical opening (closing) filters. We have also improved the classical hit-or-miss operator, defined a new edge detecting operator for binary image, and the result of this new detector is completer and smoother than the edge of Canny detector.At last, we have extend the fuzzy rank morphological operators into the color space, through the definition of the edge detector in classical mathematical morphology theory, defined the edge detector for color images in HSV color model space. After a lot of experiments, we find it could be used in extracting edges from color images.