Currently, the identification and enumeration of Giardia Iamblia cysts are based upon microscopic methods requiring individuals proficient in this area. It is a tedious process which consumes time that could be constructively used elsewhere. This project attempts to alleviate that burden by employing a computer to automatically process Indirect Fluorescent Antibody (IFA) prepared slides using digital image processing techniques. A computer controlled frame grabber, in conjunction with a CCD TV camera mounted on the epi-fluorescence microscope phototube, captures the light intensities of the objects in view under the microscope objective. The captured image is stored as pixels, with each pixel having a numerical value that can be altered using linear contrast enhancement and bit-slicing to emphasize the cysts and eliminate the majority of unwanted objects from the image. The altered image is then analyzed by a vector trace routine for typical area and perimeters characteristic to Giardia lamblia cysts. Objects in the image matching these characteristics are most likely cysts and are added to a running tally of the number of cysts present on the slide.

Included in

Biomedical Commons