ITERACINIS OPTIMIZAVIMU PAGRĮSTO AKIES DUGNO KRAUJAGYSLIŲ TRASAVIMO METODO VARIANTAS
ITERATIVE VARIANT OF THE OPTIMISATION-BASED EYE FUNDUS BLOOD VESSEL

Summary/Abstract: Eye fundus is the place where human microvasculature can be observed directly. The vascular changes that can be detected there make it possible to diagnose complications of various systemic diseases. Thus a lot of attention is paid to blood vessel detection in eye fundus images. Most of the currently used blood vessels tracking methods start the tracking either from a set of automatically generated seed points, or from the manually marked seed points. In this study, we tried to explore the possibility to start the tracing from one point that belongs to the blood vessel with a high probability. Using the method described here blood vessel tracking starts from fi nding a seed point according to the difference of red and green RGB components. The next points of the blood vessel are found by searching for the maximal values of the goal function. The study has corroborated that it is possible to achieve a blood vessel detection quality comparable to the quality achieved by other blood vessel detection methods by starting blood vessel tracking from a single point according to the described method.

• Details
• Contents