Date of Publication :5th April 2025

Abstract: This paper focuses on extracting the geometric, texture, and color numerically of red blood cells (RBCs). Several image processing processes were to segment and isolate 1000 individual RBC from the Cytoplasm. The isolation individual of blood cells was classified using three sets of integrated features. The first set, which only included geometrical features, was used to determine whether the tested blood cells were non-spherical or had a spherical shape. The types of spherical and non-spherical cells were identified using the second set, which mostly consisted of textural characteristics. The thread set, consisting the color of cells. Support vector machines (SVMs) as classification algorithm was apply to tested the features. It achieves 98% classification accuracy. The performance of a classification algorithm gives a Precision 98%, recall 97%, and Kappa statistic 0.968.

Reference :