Improving morphological image processing based on fuzzy set theory
Abstract
This article proposes the use of fuzzy set theory to improve morphological approaches in preprocessing images, particularly medical mammography images. The limitations of classical binary morphological operations reduce efficiency in grayscale images. Therefore, fuzzy erosion, fuzzy dilation, opening, and closing operations were generalized using brightness functions defined in the [0,1] range. The proposed approach demonstrated the possibility of reducing noise in image segmentation, clarifying boundaries, and improving visual quality. Additionally, through combination with Butterworth filter and histogram equalization methods, experimental results showed increased image clarity and contrast. Fuzzy morphological operations can serve as an effective tool in analyzing mammography images and as a preliminary stage in cancer diagnosis. The research results propose a new method that can be practically applied in medical image processing, segmentation, and improving their visual quality.