Astute Segmentation and Classification of leucocytes in blood microscopic smear images using titivated K-means clustering and robust SVM techniques
C Ganesh Varma, P. Nagaraj, V. Muneeswaran, K. Muthamil Sudar, M Mokshagni, Mandhala Jaswanth
Abstract
The microscopic smear images of blood are the images obtained through blood tests. So, by obtaining these images through blood tests, we can be identifying the number of diseases. As these image samples be will be in a huge number making the manual segmentation process will be more difficult and resulting in error. So here we have giving a solution of automatic segmentation for those huge sample sizes and resulting in the minimization of errors. As we know that the blood consists of three types of blood cells which are RBC that is Red blood cells, WBC means White blood cells and the platelets. So, in this category of blood cells White blood cells (WBC) are the important blood cells as they are the disease defenders and they give us a lot of scope for developing a new variety of technologies for detection and the classification of these types of blood cells i.e., WBC's. So, in our work we are using computer-based methods for the segmentation and classification process. In the line of methodology, the process involved are first the data is to be preprocessed and then the processed data is to be segmented followed by the extraction of features of the data and at the end the data is classified into 5 categories of WBC's that are easonophil, monocyte, nuetrophil, basinophil, lymphocyte. The accuracy results of these automatic segmentation and the classification process will be comparatively high to that of the manual segmentation process.