An integrated approach for Diabetic Retinopathy exudate segmentation by using Genetic Algorithm and Switching Median Filter

Diabetic Retinopathy is one of the main reason of vision damage occurred by complication of diabetes. Early detection of Diabetic retinopathy is very important for saving vision pairment and for effective treatment. Exudates are the main symptoms of Diabetic Retinopathy. Segmentation of exudates in retinal human images by using Ant Colony Optimization (ACO) technique results better than all other existing methods. But the speed of ACO is quite slow so, it becomes time consuming. The effect of noise in existing techniques is also neglected in fundus image segmentation. Therefore, in order to overcome the above stated issues an integrated algorithm is proposed. In this paper, the proposed technique firstly applied switching median filter to remove the effect of high density noise in retinal images then genetic algorithm will come in action to locate exudates in these images. The experimental results have clearly shown that the proposed technique outperforms over the available techniques in terms of sensitivity, accuracy and error rate.