Analysis of blood smear images using dark contrast algorithm and morphological filters

Abstract

In recent years, Biomedical Imaging has emerged as an effective tool in diagnosis of various diseases. In order to perform anatomy or histology of cells, Blood Smear Images are used. To process these images, enhancement plays a major role in order to increase visual quality of the image and for accurate segmentation of Region of Interest (ROI). The motive of this work is to perform enhancement using the Dark Contrast Algorithm (DCA) since it increases the intensity of darker regions, which in case of Blood Smear Images are nucleus. Further, the quality of enhanced image is evaluated using suitable Image Quality Assessment (IQA) metric. This enhanced image is segmented using Morphological Filters with appropriate structuring element to extract ROI which is nucleus and cell periphery. This helps to identify irregularities in cell periphery to detect various blood disorders. The performance of segmentation technique is assessed using Jaccard Coefficient (JC).