Search

Eyes are the organs of the visual system. The eyes are affected by eye diseases causing loss of vision. Retinitis Pigmentosa is a genetic condition which is passed down to families and it varies from person to person. Segmentation (Automatic) of external retinal layers is challenging because of the resolution of the human eye. This paper proposes a technique to classify the external retinal layers by means of KNN. It helps doctors to plan on further treatment. To segment the external layers in the retina by applying different preprocessing steps. To detect the region of interest(outer retina) the pre-processing step is performed. Second, few Framework parameters of the outer retina are calculated by fitting the candidate models. To select the Framework supported by the data Framework resolution technique is required by model selection. The detected layers are assigned by the labels (layer identification). Finally, classification by (KNN) to decide whether or not the eye's is in normal or in abnormal condition.

Retinal fundus image morphology indicates diseases like diabetic retinopathy and glaucoma like diseases. Blood vessel features of the retinal image allow ophthalmologist to perform retinal disease evaluation. Presence of lesions in the fundus image is initial sign of diabetic retinopathy. Preprocessing technique is mainly used to remove unwanted noises and it enhances some image features. Different types of pre-processing techniques are used for image processing. This paper represents the comparison of different methods used for pre-processing in retinal fundus image.