Deep Learning Approach to Predict Blood Cell and Counting

Abstract

The counting of blood cells is an important test that aids in the diagnosis of certain disorders. The technique of manually counting abnormal blood cells by an experienced specialist is exceedingly arduous, time-consuming, and imprecise, with a substantial risk of error. Automated detection of blood cells using image processing techniques is gaining prominence as a result of recent advancements. In this research, the authors offer a deep learning strategy for automatically identifying and counting three types of blood cells using the ‘you only look once' (YOLO) object detection and classification algorithm in this paper. The YOLO framework has been taught to automatically recognize and count red blood cells, white blood cells, and platelets using a modified configuration BCCD Dataset of blood smear image. Overall, our automated blood cells counting system is fast and more efficient to detect blood cells.