glaucoma is leading chronic eye dieses in the world that leads to vision lost. The main cause of Glaucoma is intrinsic deterioration of the optic nerve which leads high intraocular pressure of the eye. Manually detection of glaucoma is tedious and costly. In our work we are providing automated system for glaucoma detection which is based on fully connected conditional random filed (FC-CRF) model, it works on long and thin structure. Conditional random filed provide a platform for structure prediction. Taking benefit of current results, validating assumption and parameters of our system learned automatically with the help of structured output support vector machine. Our system trained both quantitatively and qualitatively on publically existing data sets: DRIVE, STARE, CHASEDB1 and HRF. Once we obtain segmentation results further classification is done by SVM and K-NN classifier results of our proposed system is analyzed with gold standard labeling provided each data sets in terms of TP,TN,FP and FN. importance of our proposed system is it works for enlarge structure which can provide a platform to other biomedical and biological applications.