Diabetes isrising at an alarming rate, leading to complications in eyes, kidneys, and nervous system. One of the complications in the eyes is retinopathy, which can lead to blindness in worst case. An automated tool will be of great help to ophthalmologists in the initial screening. The detection of blood vessels is very important in the diagnosis of diabetic retinopathy through image processing approach. In this paper, various techniques for blood vessel segmentation to detect diabetic retinopathy are presented and best is identified. The results obtained from this method for accuracy, specificity and sensitivity for blood vessel detection were 96.42%, 99% and 64.47% respectively.