SIFT detector based color retinal fundus image segmentation

Computerized segmentation of retinal vessels can help doctors discover diabetic retinopathy, Retinal Tear and Detachment. Because the anatomy is complicated changing, automated vessel still a difficult task. The Expectation-Maximization approach for vascular proposed in this paper, which uses matched filter (MF) that combines Mask based Blood Vessels Extraction with global image threshold to extract network. Firstly, Contrast Limited Adaptive Histogram Equalization method applied order improve contrast between blood background. Scale Invariant Feature Transform Detector used estimate key point. expectation maximization technique vessels, mask-based extraction picture threshold. evaluation was assessed utilizing ANN evaluate performance Normal Abnormal Images. algorithm compared analyzed using other approaches.