Feminna Sheeba, Robinson Thamburaj, Joy J. Mammen, Hannah M. Thomas Thevarthundiyil, and Atulya K. Nagar
Segmentation, watershed algorithm, morphological operations, discrete wavelet transformation, reversible watermarking, image authentication
This paper proposes an automated method of segmentation of white blood cells, in order to calculate the differential count and watermarking of the results in peripheral blood smear images. The cytoplasm and the nucleus of the WBCs are segmented separately. Each cell is then tested for a cytoplasm-nucleus combination, which is the indication of a WBC's existence. The segmented cytoplasm and nucleus are then analyzed to classify the WBCs using active shape models programmed to represent the various classes of WBCs.
Important Links:
Go Back
