Vessel Segmentation In Retinal Image Based On Urvelet Transform And Morphology

The retina is the innermost layer of the eye, it is composed of many importantphysiological structures. Vessels are the most important structures. Many diseases canchange the shapes and the structures of the blood vessels, so the analysis for retinal vesselscan prevent some disease from occurring and diagnose some early diseases. Conventionalmanual segmentation for retinal images is a time-consuming and trivial task. Intrinsiccharacteristics of retinal images make the bloods vessels detection process difficult. Thereare many good methods to detect blood vessels in retinal image proposed by researchers.But most of them trend to lose thin vessels in the low contrast. So it is urgent to make theautomatic quick segmentation for retinal images with computer.. Due to the high ability of the Curvelet transform in representing the edges, wepropose a method to enhance the retinal image contrast. The method is that modificationsof Curvelet transform coefficients to enhance the retinal image edges. We introduce anonlinear function to modify Curvelet transform coefficients in such a way that details ofthe small amplitude are enlarged.Definition of the function parameters based on somestatistical features of Curvelet coefficients of the input image. The methods can enhancethe retinal image contrast. Afterward, we use modified top-hat to detect edges andmorphological operators by reconstruction to eliminate the false-edges not belonging tothe vessel tree while trying to preserve the thin vessels unchanged. We use multi-structureelements because the directionality feature of them makes it an effective tool in edgedetection. Finally, we use the threshold segmentation based on entropy to detect vessels.The performance of the proposed method is evaluated on the available DRIVEdatabases, and the method can detect all the big vessels and many small thin vessels, andthe runtime of our algorithm is below50s. So our proposed method is a quite fast indetecting the vessels with a reliable accuracy.