Segmentation of Biological Cell Images for Sonification

Cervical cancer is one of the most preventable/treatable forms of cancer, due to the fact that precursor signs for the disease can be detected in microscopic examination of cervical cells. Currently these examinations are purely visual, but the project reported aims to process the visual images to present them in a complementary auditory form and thereby to improve diagnostic accuracy. Standard mean-shift image processing techniques have been successfully applied to extract nuclear data suitable for presentation in an auditory form.