ABSTRACT
Individuals with cancer have a high fatality rate. To fight against cancer, early diagnosis is essential. Leukemia is cancer of the body's blood-forming tissues. Many types exist, such as acute lymphoblastic leukemia, acute myeloid leukemia and chronic lymphocytic leukemia. Most of the affected people are not aware they have leukemia since it does not have any indications in the body. This project proposes a method to predict acute lymphoblastic leukemia with an image of microscopic blood cells. The proposed model has a better image processing technique where feature extraction is done by segmentation algorithm of watershed and k-means clustering for colored blood cell images. Classification is done using support vector machines. The proposed system will accept the blood image as input and predict whether it is infected or not. This method has been utilized on various images of the cancerous cell and it consistently gives the correct output. This method improves accuracy with minimum time.
