This study proposes a strategy for detecting dark and bright spots in the retinal images which can lead to earlier detection and better treatment of the retinal pathology.

Nowadays, diabetes is considered to be one of the highly prevalent diseases all over the world. Many techniques were proposed for early detection of the abnormalities through black and white spots in the retinal images, help the ophthalmologists to recognize the retinopathy earlier. A computer based method is applied through an algorithm. Then the results obtained in the previous step were analyzed. Our presented model of the computer based Diabetic Retinopathy System make benefit from a 3 staged procedure. The first phase includes the removal of the parts of the input image that are of law quality. In the second phase the blood vessels are enhanced and segmented and then OD is localized and process of detection is done. In the final phase, the dark and bright spots were detected.

Early diagnosis of retinopathy in persons with diabetes is critical in preventing blindness and visual impairment in adults. The results of this study show high accuracy and precision in comparison with manual segmentation performed by two different experts, and is free of many problems that are observed in the other methods especially than manual methods (P<0.05).

Applying the novel techniques of the processing can make the task easier and we can easily perform the process of detecting dark and bright spots of the retinal images as a symptom of DR. The information obtained from image processing is very efficient. In a short time a large volume of images of the retina can be checked. Automatic image processing also significantly reduced the rate of medical errors.