Extraction of Multiple Cellular Objects in HEp-2 Images using LS Segmentation
P. Suhail Parvaze, S. Ramakrishnan · IEIE Transactions on Smart Processing and Computing · 2017
In this work, an attempt has been made to extract multiple objects from Human Epithelial Type 2 (HEp-2) cell images using LS based segmentation. Thirteen Positive and intermediate intensity level images that consist of Homogenous, Centromere and Nucleolar patterns are considered in this study. The results show that LS based segmentation is able to extract objects of interest from the images. The extraction efficiency is found to be better for LS segmented images. Among the considered cell shapes, Centromere pattern is found to be sensitive for the extraction method. The overall cell extraction accuracy is found to be 90% in both positive and intermediate intensity images. It appears that this approach with LS for segmenting objects and differentiating positive and intermediate images is found to be useful for automated analysis of autoimmune diseases.