Morphological Contour Based Blood Vessel Segmentation in Retinal Images Using Otsu Thresholding

2. Previous Related Work

Retinal blood vessel segmentation becomes an essential step for diagnosis of various retinal disorders. There exist several vessel extraction techniques which either extract the blood vessels directly or enhance the blood vessels and then extract the vessels. In addition to the blood vessels in the image, there are several other components such as the fovea which is the darker region at the middle of the image and the optical disc which is at the intersection of the largest red region and the brightest region of the image. Rahul et al, (2016) proposed a method based on morphological operations for blood vessel extraction. In this work the preprocessed image undergoes morphological operations followed by feature extraction using Gabor filter. Final vessel structure is segmented based on the Euclidean distance. Khan et al, (2016) proposed a method based on eigen values for retinal blood vessel extraction. The proposed method applies eigen based approach at different scales to segment thick and thin blood vessels. Finally, the pixels are classified based on Otsu thresholding as either vessels or non-vessels. Toufique et al. (2016) contrast sensitive approaches are integrated with conventional algorithms which works well for different images conditions thereby improving the sensitivity achieved by the retinal vessel extraction technique (Soomro, Gao, Khan, Khan, & Paul, 2016).