Morphological Image Processing and Blob Analysis for Red Blood Corpuscles Segmentation and Counting

Abstract

this paper presents a proposed procedure using a morphological image processing on a 960x720 pixels blood sample image. The principles of hemocytometer counting method were also performed to count and compare the output readings from the laboratory results. The red blood corpuscles (RBC) count is one of the essential elements that medical practitioners used for medical diagnosis of patients. Because of the morphological features of the RBC on the image file, different approach yields in the processing including image enhancement to reveal certain features such as edges and contours in the RBC. The concavity in the RBC was useful in the blob analysis with the watershed algorithm that leads to the segmentation of overlapping cells in clusters. The proposed procedure gives 96.042% accuracy for female test subject and 95.559% on the male test subject using two trials each with ten samples on each trial. The proposed procedure is successful by getting a results compared to the manual count performed in the laboratory. The use of the application program created using MatLab for blob analysis yield a good results in recognizing red cells and for counting each segmented cells.