Evaluation of retinal vessel segmentation methodologies based on combination of vessel centerlines and morphological processing

The categorization of retinal vessels morphological features and the investigation of their branching patterns are used in the process for automated diagnosis and screening of ophthalmologic diseases. Mathematical morphology has been materialized as a proficient technique for quantifying the retinal vasculature in ocular fundus images. In this paper, the performance comparison of two retinal vessel segmentation approaches based on combination of multi scale morphological reconstruction and morphological bit plane slicing with the vessel centerlines is presented. The segmentation accuracy and the processing time are taken as the performance criteria. These approaches are tested on two publicly available databases and results demonstrate that morphological bit plane slicing outperforms other approach in respect of processing time without a significant degradation of sensitivity and specificity.