Identification of human blood cells

The objectives of the mini-project is to identify the three types of blood cells present in human blood, namely red blood cells, white blood cells and blood platelets. Also, the two main types of white blood cells (phagocytes and lymphocytes) need to be differenciated. These cells are identified from a PGM image and the number of each type of cells are counted. Other (abnormal) cells present in the image need are also detected, though not identified.

Techniques Used

 • Median filtering
 • Image thresholding
 • Mathematical opening (erosion and dilation)
 • Image Labeling
 • Classification (area and circularity)

(Application developed in Visual C++)

Software Screenshots


The red blood cells in the above example are labeled as R, the lymphocytes as L, the phagocytes as H, the platelets as P. Cells touching the image borders are not taken into consideration. Cells that are counted are outlined in black.

Download the report (PDF file, 779 KB)