An automatic system for the blood vessel segmentation in retinal, on the basis of colour and texture components, is presented in this paper. The key utility of this paper is to be a useful tool for medical disciplines to detect possible cases of diabetic retinopathy instrument. In this paper, we will focus on the extraction of retinal blood vessel. A novel approach using differential evaluation (DE) on OTSU’s algorithm for blood vessel segmentation, has been presented in this article. Finally, statistical calculations are performed to quantify the Accuracy, sensitivity and robustness of the proposed method, after performing a battery of tests on various retinographies belonging to two different databases, STARE database and DRIVE database, the results being certified specialist beforehand.
Volume 11 | Issue 10