Automated Thalassemia cell image segmentation using hybrid Fuzzy C-Means and K-Means

Thalassemia is a form of hereditary disease. Thalassemia is one of the world's most common illnesses. The morphology of red blood cells is most affected by this disorder. This research proposes a new method of automatically segmenting red blood cells from microscopic blood smear images. The research suggests a novel combination of image processing techniques and extensive preprocessing to achieve superior segmentation performance. In this work, the eleven designated color spaces, with six filters and three contrasts enhancing, Fuzzy c-means and K-means segmentation studied using five evaluation parameters. This evaluation is based on the ground truth image. The Photoshop program performs novel ground truth techniques for multi-object sense (RBC cells). The optimization of all image processing stages was obtained through local image datasets (258 images) obtained from seven thalassemia patients in the Erbil – thalassemia center and five samples of normal blood cells in Children Raparin Teaching Hospital. The image was captured with different light intensities (low, medium, high) and with /without a yellow filter in Biophysics Research lab /Education College / Salahaddin University –Erbil. This study found that the best light intensity for image slide capture utilizing a microscope was medium without using a yellow filter with an accuracy of 0.91± 0.14 and a performance of 95.34%.