Publication

Abstract : Counting a lymphocytes is bit a time consuming and tedious task and also not that interesting but we should not neglect the task because doctor should get an accurate report about a patient so that it will help to understand a doctor more about a patient and sometimes the manual counting of lymphocytes may also lead to improper / incorrect count of a lymphocyte which affect in analyzing a patient so we should not neglect the task. Because of these disadvantages we are developing an automated tool to count the number of lymphocytes using image processing. Initially image is converted into gray scale to get the maximum accuracy in the result, and then the image is compressed so that it can be stored in a minimal storage space, then edge detection of a cell is done through canny detection process to extract the counter boundary of the blood cell, then dilation and erode is applied to enlarge the interested cell and contract the non-interested cells by using a morphological characteristics ,then watershed algorithm is applied to segment the cytoplasm and nucleus from the blood cell. Finally counting of lymphocytes is performed.