A New Seeded Region Growing Technique for Retinal Blood Vessels Extraction

摘要: Distribution of retinal blood vessels (RBVs) in images has an important role the prevention, diagnosis, monitoring and treatment diseases, such as diabetes, high pressure, or heart disease. Therefore, detection exact location RBVs is very for Ophthalmologists. One frequently used techniques extraction these is region growing-based Segmentation. In this paper, we propose a new growing (RG) technique extraction, called cellular automata‑based segmentation. RG often require manually seed point selection, that is, human intervention. However, due to complex structure images, manual tracking them very difficult. make our proposed full automatic, use automatic selection method. The was tested on Digital Retinal Images for Vessel Extraction database three different initial sets and evaluated against segmentation images available at database. Three quantitative criteria including accuracy, true positive rate false rate, were considered to evaluate visual scrutiny results show that, using automata for extracting promising. here correct seeds have effective role final segmentation.