Diabetic Retinopathy is a complication in which the retina of the eye gets damaged due to secretion of fluid from blood vessels to retina, It may also cause vision loss. It occurs due to damage of blood vessels in the retina. In this approach, we suggest an ANN method to detect the presence of unnatural new blood vessels. This new concept is used for distinguishing age-related problems i.e., age-related macular degeneration (AMD). Our new improved matched filters distinguishes the higher efficiency in true up positive feedback and less efficiency in false observation in previously used filter based on vessel separation. So the aim is to provide an automated, easy to use, UI based detection system preparing the raw images to be processed by applying general operations such as cropping and resizing the images and applying noise reduction filters, finding global mean and using it for thresholding and also applying a pad for distinct borders to avoid confusion. The DR can be observed at rapid levels simply and destruction of the retina will be less. The key research methodology is stated that describing patterns that report the characteristics of an entire image such as outcome arranged images are graceful with dissimilar filters, geomorphologic tools, noise contraction, background separate and vessel separation. Comparison of ANN technique, and support vector machine (SVM) technique used for the classification and grading technique based on supervised learning.