Identification of Trypanosoma with digital image processing

The present work shows a segmentation methodology of images for the identification of Trypanosoma cruzi parasite at adult stage with techniques of digital image processing. To implement the method, 83 sample images of blood smears stained with Giemsa dye containing the parasite were used. The identification method uses a segmentation in the HSI color space (Hue, Saturation, Intensity), which is separately applied to main components in the image, such as background, surrounding cells and the parasitic. The identification is achieved by combining each sub-segmentation result and eliminating any noise after performing these sub-segmentations.