Exudates Detection in Fundus Image Based on Fuzzy C-Means Clustering and Edge-Aware Model

doi:10.3778/j.issn.1002-8331.1904-0388

Computer Engineering and Applications ›› 2020, Vol. 56 ›› Issue (11): 192-199.DOI: 10.3778/j.issn.1002-8331.1904-0388

Previous Articles Next Articles

Exudates Detection in Fundus Image Based on Fuzzy C-Means Clustering and Edge-Aware Model

LIU Juntao, WANG Sujuan, LIN Xiaoming, LIU Zuoshi, TAN Jianhui, SONG Dan

1.College of Mechanical and Electrical Engineering, Jiangxi University of Science and Technology, Ganzhou, Jiangxi 341000, China
2.Guangdong Shunde Innovation Design Institute, Shunde, Guangdong 528311, China
3.College of Automation, Guangdong University of Technology, Guangzhou 510006, China
4.College of Information Science and Engineering , Guilin University of Technology, Guilin, Guangxi 541006, China

Online:2020-06-01 Published:2020-06-01

结合FCM聚类和边缘感知模型的眼底渗出物检测

刘俊涛，王素娟，林晓明，刘祚时，谭俭辉，宋丹

1.江西理工大学机电工程学院，江西赣州 341000
2.广东顺德创新设计研究院，广东顺德 528311
3.广东工业大学自动化学院，广州 510006
4.桂林理工大学信息科学与工程学院，广西桂林 541006

Abstract

Abstract:

Exudates in the fundus image is one of the early symptoms of Diabetic Retinopathy（DR）, a method for detecting exudates by combining Fuzzy C-Means（FCM） clustering and edge-aware model is proposed. In order to ensure the accuracy and efficiency of post-detection, the fundus image is firstly subjected to preprocessing such as enhanced contrast and equalized brightness, and then the exudates candidate region is segmented by FCM clustering, after that the candidate is determined by the edge-aware model based on the judgment of neighborhood grayscale difference. Finally, the real exudates area is obtained by removing the optic disc area. The approach is evaluated on the public fundus image data set, the sensitivity of the algorithm is 86.65%, the specificity is 94.79%, the positive predictive value is 95.14%, and the accuracy is 92.09%. The results show that the method can effectively realize the automatic detection of the fundus exudates.

Key words: exudates detection, image preprocessing, fuzzy C-means clustering, edge-aware model

摘要：

眼底图像中渗出物是构成糖尿病视网膜病变（Diabetic Retinopathy，DR）的早期症状之一，提出一种结合模糊C-均值（Fuzzy C-Means，FCM）聚类和边缘感知模型的方法实现对渗出物的检测。为保证后期检测精度和效率，对眼底图像进行增强对比度和均衡亮度等预处理操作，用FCM聚类分割出渗出物候选区域，利用基于判断邻域灰度差异的边缘感知模型对候选区域进行筛选，通过移除视盘区域，从而得到真实的渗出物区域。在公开的数据集上进行实验，算法的灵敏度为86.65%，特异性为94.79%，阳性预测值为95.14%，准确度为92.09%。结果表明，该方法能够有效实现对眼底渗出物的自动检测。

关键词: 渗出物检测, 图像预处理, 模糊C均值聚类, 边缘感知模型

LIU Juntao, WANG Sujuan, LIN Xiaoming, LIU Zuoshi, TAN Jianhui, SONG Dan. Exudates Detection in Fundus Image Based on Fuzzy C-Means Clustering and Edge-Aware Model[J]. Computer Engineering and Applications, 2020, 56(11): 192-199.

刘俊涛，王素娟，林晓明，刘祚时，谭俭辉，宋丹. 结合FCM聚类和边缘感知模型的眼底渗出物检测[J]. 计算机工程与应用, 2020, 56(11): 192-199.

[1]	PENG Xiaohong, LIANG Zixiang, ZHANG Jun, CHEN Rongfa. Review of Underwater Image Preprocessing Based on Deep Learning [J]. Computer Engineering and Applications, 2021, 57(13): 43-54.
[2]	CHAO Jingjing, SHEN Wenzhong, SONG Tianshu. Eye Location Algorithm of Binocular Iris Image Based on HOG and Cascade SVM [J]. Computer Engineering and Applications, 2019, 55(9): 184-189.
[3]	LIU Qiang1, SHI Hong1, WANG Pingxin2，3, YANG Xibei1. Three-Way Clustering Analysis Based on [ε] Neighborhood [J]. Computer Engineering and Applications, 2019, 55(6): 140-144.
[4]	JIA Wei1，2, HUA Qingyi1, ZHANG Minjun1, CHEN Rui1, JI Xiang1, WANG Bo1，3. Semi-supervised classification of mobile interface pattern using improved extreme learning machine [J]. Computer Engineering and Applications, 2018, 54(2): 11-19.
[5]	WANG Wenhui1, YANG Geng1, GE Wei1, LIU Peidong2, QIAN Chen3. Fuzzy C-Means clustering algorithm under noise uneven condition and its application [J]. Computer Engineering and Applications, 2018, 54(19): 172-178.
[6]	LIU Cangsheng, XU Qinglin. Fuzzy C-means clustering algorithm based on density peak value optimization [J]. Computer Engineering and Applications, 2018, 54(14): 153-157.
[7]	FENG Yufang, LU Houqing, YIN Hong. Fuzzy C-means clustering image segmentation method based on multi-chain quantum bee colony algorithm [J]. Computer Engineering and Applications, 2017, 53(24): 8-14.
[8]	HOU Xiaofan, WU Chengmao. Fast KFCM clustering segmentation algorithm with local information [J]. Computer Engineering and Applications, 2017, 53(13): 189-195.
[9]	EN De, ZHANG Fenglei, ZHANG Zhao, HU Shengqiang. Application of fuzzy entropy in speech endpoint detection in car environments [J]. Computer Engineering and Applications, 2016, 52(10): 147-150.
[10]	DUAN Baobin1，2, HAN Lixin2, XIE Jin1. Fuzzy C-means clustering algorithm based on stacked sparse autoencoders [J]. Computer Engineering and Applications, 2015, 51(4): 154-157.
[11]	YU Yong1, ZHANG Yunwei1, WANG Jing2, WANG Dalong2, WANG Yanjun2, BAO Jun1. Research on recognition of tobacco leaf disease based on computer vision [J]. Computer Engineering and Applications, 2015, 51(20): 167-171.
[12]	WANG Junling1、2, WANG Shitong1, BAO Fang2, ZHOU Jianlin3. Fast and robust fuzzy C-means clustering algorithm based on space distance of nearest neighbors [J]. Computer Engineering and Applications, 2015, 51(1): 177-183.
[13]	ZENG Minghua1，2, WANG Yinsong1, YANG Xiaoguang1. Study on index system for evaluating supplier collaboration capabilities [J]. Computer Engineering and Applications, 2014, 50(4): 18-23.
[14]	ZHANG Fei, FAN Hong. Research on medical image segmentation based on fuzzy C-means clustering algorithm [J]. Computer Engineering and Applications, 2014, 50(4): 144-151.
[15]	LIN Mugang1, LIU Fangju2, TONG Xiaojiao1. Fuzzy clustering algorithm based on firefly algorithm [J]. Computer Engineering and Applications, 2014, 50(21): 35-38.

Exudates Detection in Fundus Image Based on Fuzzy C-Means Clustering and Edge-Aware Model

结合FCM聚类和边缘感知模型的眼底渗出物检测

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics