Retinal images are widely used for diagnostic purposes by ophthalmologists. Therefore, these images are suitable for digital image analysis for their visual enhancement and pathological risk or damage detection. Here, we discuss two different approaches towards optic disc segmentation. We review and analyze a prior work based on a pixel-wise contour extraction for optic nerve head segmentation along with the effects of processing retinal images under lossy compression. This analysis revealed the need for a more general and robust approach. Finally, we introduce a different strategy for optic disc segmentation based on the use of active contours and color mathematical morphology.
展开▼
