A Study on the Assesment of Leukocyte with Hybrid Imaging Technique

Abstract

In recent years, a considerable number of bio-image and assisted disease detection systems are proposed and implemented in medical clinics to identify the diseases. The proposed work aims to implement an imaging approach to segment the Leukocyte (White-Blood-Cell) section from the thin blood smear image. The Leukocyte segmentation and evaluation is an initial process and it is related with the blood-sample analysis. The proposed technique implements the integration of RGB image threshold and segmentation technique to extract the White-Blood-Cell (WBC) section with better accuracy. In this work, the Bat-Algorithm (BA) and Shannon's entropy assisted thresolding is initially executed to enhance the WBC by eliminating the other sections and later the Watershed-Algorithm (WA) is executed to extract the WBC. Finally, a relative assessment with the extracted WBC and the ground-truth is implemented and based on the accuracy; the performance of the proposed technique is authenticated.