| Summary: | Glaucoma is one among major causes of blindness in working population. Early detection of Glaucoma through automated retinal image analysis helps in preventing vision loss. Optic disk (OD) segmentation from retinal images is the preliminary step in developing the diagnostic tool for early Glaucoma detection. In this paper, we have presented a novel hierarchical technique for fast and accurate OD localization and segmentation. Retinal vasculature and pathologies are delineated and removed by using morphological operations at preprocessing stage followed by circular Hough transform for OD localization. The precise boundary of OD is obtained by calculating the region of interest and applying a novel polar transform-based adaptive thresholding. The methodology is evaluated on a number of publicly available retinal image sets, which includes MESSIDOR, DIARETDB1, DRIONS-DB, HRF, DRIVE, and RIM-ONE, and it has shown considerable improvement over existing methods in terms of accuracy and processing time.
|
|---|
