In recent days, retinal diseases are gradually increasing and results in blurred vision, side vision defects and in a worst situation, people may lost their vision. From the literature, it is noticed that the macular degeneration, diabetic eye disease and Glaucoma are termed as the important research domain. With that motivation, this research is mainly focused on Glaucoma diseases and its detection. Glaucoma segmentation is critically significant for medical research and diagnosis. The conventional research is mainly focused on developing the retinal image segmentation process and detection. There are various other possibilities to merge previous algorithms to form best hybrid segmentation process. Likewise, the research is focused on improving the average correlation, average F-score and reduce the average boundary distance of the segmentation process. For achieving such objectives, the adaptively regularized Kernel is considered for maintaining the robustness and maintain the image quality. Fuzzy C-Means clustering is considered for deploying effective clustering and minimize the computation cost. The proposed modified Kernel Fuzzy C-Means (MK-FCM)segmentation method has achieved a maximum average F-score of 0.975, average boundary distance of 10.112 pixels and average correlation coefficient of 0.916.