.png){kind=icon}
返回
Retinal vessel segmentation using multifractal characterization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2020.106439.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
This study presents a supervised classification method for the segmentation of retinal vessels using fundus images. This work proposes a novel retinal vasculature segmentation method based on multifractal characterization of the vessels to minimize the noise and enhance the vessels during segmentation. The Holder exponent, a multifractal measure is employed for the first time to segment the retinal vessels. The Holder exponent is used to quantify the local regularity of the vessels. The Holder exponents are computed from the Gabor wavelet responses for the effective segmentation of vessels, which is a novel feature of the method. The Gaussian mixture model (GMM) classifier is used for the classification of pixels. Different multifractal measures used to compute the Holder exponents are evaluated for the output quality. The effectiveness of the method is evaluated using three publicly available datasets for fundus images namely, DRIVE, STARE and CHASE_DB1. The proposed method provides robust segmentation of retinal vessels for both normal and abnormal images (i.e., images with pathologies) at a reasonable segmentation speed and the method is also simple to configure. It achieves an average accuracy and area under the receiver operating characteristic curve of 0.948 and 0.959 respectively on the DRIVE dataset; 0.9542 and 0.9711 respectively on the STARE dataset; 0.9459 and 0.9592 respectively on the CHASE_DB1 dataset; 0.9500 and 0.9623 respectively on the abnormal images. (C) 2020 Elsevier B.V. All rights reserved.
Keyword:
Multifractal
Holder exponent
Fundus
Retinal vessel segmentation
Gabor wavelet
Morphological reconstruction
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
6.6
论文数:
1.4W
被引数:
4.8W
机构
引用论文
A Test of the Catering Theory of Dividends: The case of the Japanese Electric Appliances Industry
OAAI
OAAI分享收藏Automated Vessel Segmentation Using Infinite Perimeter Active Contour Model with Hybrid Region Information with Application to Retinal Images使用具有混合区域信息的无限周长活动轮廓模型的自动血管分割及其在视网膜图像中的应用OAAI
OAAI分享收藏Trainable COSFIRE filters for vessel delineation with application to retinal images.png)
MEDICAL IMAGE ANALYSISOAAI
MEDICAL IMAGE ANALYSISIF11.8
OAAI分享收藏A multimodal registration algorithm of eye fundus images using vessels detection and Hough transform
PREAI
分享收藏Segmentation of retinal blood vessels by combining the detection of centerlines and morphological reconstruction结合中心线检测和形态学重建的视网膜血管分割
PREAI
分享收藏Retinal vessel delineation using a brain-inspired wavelet transform and random forestPATTERN RECOGNITION
PATTERN RECOGNITIONIF7.6
PREAI
分享收藏
查看更多内容