Optic Disc Detection

This code detects blood vessels and optic discs in retinal images. Effective optic disc detection in color retinal imagery serves as a fundamental step in automated retinal image analysis systems. The proposed methodology implements a novel approach for automatic optic disc localization and precise boundary detection. The algorithm utilizes iterative thresholding followed by connected component analysis to approximate the optic disc center, then applies a geometric model based on implicit active contour models to define exact boundaries. Validation on a curated database of 148 retinal images demonstrates 99.3% localization accuracy, with boundary detection achieving sensitivity of 90.67±5% and specificity of 94.06±5% compared to human expert annotations.