SEGMENTATION OF IMAGES OF EYE GROUND BLOOD VESSELS INVOLVING APPLICATION OF FUZZY IMAGING

Segmentation of images is an important task while processing images. Among the most widespread methods are methods based on pixel clustering, histogram methods, morphological methods, watershed segmentation, multiscale segmentation, and others. A promising trend in image processing is the use of fuzzy logic methods and the fuzzy set theory. Their application makes it possible to improve the quality of processing by providing information in a fuzzy form. The article proposes a new method for images segmentation involving boundaries detection based on the fuzzy representation of the image and fuzzy pixels. The membership functions are proposed for describing fuzzy pixels, and the requirements for their form and type are provided. The most suitable membership functions for fuzzy imaging are the s-function and the π-function. A description of a new method for boundaries detection based on the Sobel operator and the developed fuzzy type of image is described. In this case, standard calculations of the image brightness gradient are supplemented with their fuzzy versions which are then combined to obtain the final result. The experimental verification of the developed method is carried out using the example of eyeground images. In addition to the fuzzy detection of boundaries for the detection of blood vessels, the images were subjected to pre-processing (halftone imaging, mask matching, contrasting), morphological operators (thinning of boundaries, dilatation), and an algorithm for removing small details was applied. During testing, the developed algorithm showed acceptable results in terms of segmentation of blood vessels. In the future, a fuzzy image model can be extended to use fuzzy features of the second and higher types