Automatic optic disc boundary extraction based on saliency object detection and modified local Gaussian distribution fitting model in retinal images

Accurate optic disc(OD) localization and segmentation are two main steps when designing automated screening systems for diabetic retinopathy. In this paper, an OD segmentation approach based on the saliency object detection and modified local Gaussian distribution fitting model(LGDF) is proposed. This approah consists of two stages: In the first stage, the saliency detection technique is introduced to the enhanced retinal image with the aim of locating the OD; in the second stage, the OD boundary is extracted by the modified LGDF model with oval-shaped constrain. The performance of proposed approach is tested on the public DIARETDB1 database. Compared to the state-of-the-art approaches, the experimental results show the superiority and effectiveness of the proposed approach. © 2019, Editorial Office of Control and Decision. All right reserved.