首页>
外文OA文献
>A Fully Convolutional Neural Network based Structured Prediction Approach Towards the Retinal Vessel Segmentation
【2h】
A Fully Convolutional Neural Network based Structured Prediction Approach Towards the Retinal Vessel Segmentation
展开▼
机译：基于完全卷积神经网络的结构预测 视网膜血管分割的探讨
展开▼
免费
页面导航
摘要
著录项
相似文献
相关主题
摘要
Automatic segmentation of retinal blood vessels from fundus images plays animportant role in the computer aided diagnosis of retinal diseases. The task ofblood vessel segmentation is challenging due to the extreme variations inmorphology of the vessels against noisy background. In this paper, we formulatethe segmentation task as a multi-label inference task and utilize the implicitadvantages of the combination of convolutional neural networks and structuredprediction. Our proposed convolutional neural network based model achievesstrong performance and significantly outperforms the state-of-the-art forautomatic retinal blood vessel segmentation on DRIVE dataset with 95.33%accuracy and 0.974 AUC score.
展开▼
