An Automatic Cognitive Graph-Based Segmentation for Detection of Blood Vessels in Retinal Images

Table 1

Illustrative study of the number of components from fully connected to locally connected layer of selected examples from the DRIVE training, DRIVE testing, ARIA, and STARE datasets.

Datasets	DRIVE training	DRIVE testing	ARIA	STARE
Initial layer	329960	329960	442368	423500
Initial layer with ROI	210702	208174	350013	288761
Spectral-layer	9874	9816	15094	12917
BTH-layer	3285	3719	3130	4968
Dijkstra-layer	2	2	2	2