Implementation of Morphological To Detect Disease Based on Image of Red Blood Cells Ali Akbar Lubis, Agnes Irene Silitonga, Adi Widarma
Universitas Negeri Medan
Abstract
Disease can be recognized visually because it has unique shape and color characteristics. The purpose of this study is to conduct research on malaria-based parasitic cell infections. In this research, there are 20 images used as malaria detection research data divided into 5 normal images and 15 malaria recovery images obtained through the dataset. The process carried out in malaria detection to input RGB image after that the image takes a grayscale image, then improve the image quality to eliminate noise that occurs in the image with the median filtering method. The results of these images are binary images. In the conversion process, the hole is carried out on the object so that the filling process is carried out. After that the cell count is done by the 4-connected method, the edge detection process uses the boundary, then the malaria collection using the color threshold and the malaria phase collection.
Keywords: image processing, disease, red blood cells, 4-connected method
