Caliber fuzzy c-means algorithm applied for retinal blood vessel detection

Retinal blood vessel detection employs a vital role in finding of retinal diseases like diabetic retinopathy and glaucoma. This paper presents an innovative unsupervised retinal blood vessel detection technique. First step is to generate a vessel enhanced image, then using calibre fuzzy c-means (CFCM) technique, first cluster the retinal image; next the clustered image is passed to the canny edge operator and finally post process the retinal image. CFCM clustering method for blood vessel detection is based on the choice of the number of clusters value. By using CFCM clustering function, compute the cluster centre, which commonly divides the image into four clusters. The proposed technique is obviously forceful into the modification of fuzzy c-means with canny algorithm. The proposed algorithm accomplishes an accuracy of about 95% of retinal images from three datasets DRIVE, STARE, and CHASE_DB1.