A HIERARCHICAL IMAGE MATTING MODEL FOR BLOOD VESSEL SEGMENTATION IN FUNDUS IMAGES

Abstract

In this paper, a hierarchical image matting model is proposed to extract blood vessels from fundus images. More specifically, a hierarchical strategy is integrated into the image matting model for blood vessel segmentation. Normally the matting models require a user specified trimap, which separates the input image into three regions: the foreground, background and unknown regions. However, creating a user specified trimap is laborious for vessel segmentation tasks. In this paper, we propose a method that first generates trimap automatically by utilizing region features of blood vessels, then applies a hierarchical image matting model to extract the vessel pixels from the unknown regions. The proposed method has low calculation time and outperforms many other state-of-art supervised and unsupervised methods. It achieves a vessel segmentation accuracy of 96:0%, 95:7% and 95:1% in an average time of 10:72s, 15:74s and 50:71s on images from three publicly available fundus image datasets DRIVE, STARE, and CHASE DB1, respectively. 

Existing System

Large kernel matting Laplacian, and achieves a fast matting algorithm

Proposed System

In this method, the trimap of an input fundus image is generated automatically by utilizing region features of blood vessels. Creating the trimap of the input fundus image automatically includes two main steps: 1) Image Segmentation and 2) Vessel Skeleton Extraction. Image segmentation aims to separate the input image into three regions: the vessel (foreground), background and unknown regions. Vessel Skeleton Extraction aims to further distinguish the unknown regions and provide more information on blood vessels. Hierarchical image matting model is proposed to label the pixels in the unknown regions as vessels or background in an incremental way. Since some non-vessel regions may still exist in the final segmented vessel image, the regions are abandoned to remove these non-vessel regions. 

Architecture

BLOCK DIAGRAM