Diabetic retinopathy is considered as the root cause of vision loss for diabetic patients. However, if the symptoms are identified earlier and a proper treatment is provided through regular screenings, blindness can be avoided. In order to lessen the cost of these screenings, modern image processing techniques are used to voluntarily detect the existence of abnormalities in the retinal images acquired during the screenings. Exudates are a major indicator of diabetic retinopathy that can possibly be quantified automatically. This paper focuses on automatic detection of diabetic retinopathy exudates in color fundus retinal images. A series of experiments on classification of hard and soft exudates is performed with the use of image processing techniques. Initially the color fundus retinal images are subjected to preprocessing for CIELab color space conversion and Fundus region detection using binarization and mathematical morphology respectively. Subsequently nonlinear diffusion segmentation is employed to encapsulate the variation in exudates and lesion boundary criteria pixels. To prevent the optic disc from interfering with exudates detection, the optic disc is detected and localized with the aid of region props and color histogram. Exudates are detected with the aid of thresholding color histogram, which is used to classify the hard and soft exudates pixel from the color fundus retinal image. Experimental evaluation on the publicly available dataset DIARETDB1 demonstrates the improved performance of the proposed method for automatic detection of Exudates. These automatically detected exudates are validated by comparing with expert ophthalmologists' hand-drawn ground-truths. Sensitivity, Specificity and Accuracy are used to evaluate overall performance. The overall sensitivity, specificity and accuracy of the proposed method are 89.78%, 99.12% and 99.07%, respectively.