计算机应用 ›› 2013, Vol. 33 ›› Issue (02): 543-566.DOI: 10.3724/SP.J.1087.2013.00543

• 多媒体处理技术 • 上一篇    下一篇

基于多尺度匹配滤波和集成学习的眼底图像微脉瘤检测

彭英辉1,2,张东波1,2,沈奔1,2   

  1. 1. 湘潭大学 信息工程学院,湖南 湘潭 411105
    2. 智能计算与信息处理教育部重点实验室(湘潭大学),湖南 湘潭 411105
  • 收稿日期:2012-08-13 修回日期:2012-09-07 出版日期:2013-02-01 发布日期:2013-02-25
  • 通讯作者: 彭英辉
  • 作者简介:彭英辉(1987-),男,湖南湘潭人,硕士研究生,主要研究方向:模式识别、图像处理;
    张东波(1973-),男,湖南隆回人,教授,博士,主要研究方向:模式识别、图像处理、神经网络集成;
    沈奔(1986-),男,湖南湘乡人,硕士研究生,主要研究方向:模式识别、图像处理。
  • 基金资助:
    湖南省科技厅资助项目;湖南省教育厅资助科研项目;湖南省重点学科建设项目

Microaneurysm detection based on multi-scale match filtering and ensemble learning

PENG Yinghui1,2,ZHANG Dongbo1,2,SHEN Ben1,2   

  1. 1. College of Information Engineering, Xiangtan University, Xiangtan Hunan 411105, China
    2. Key Laboratory of Intelligent Computing and Information Processing, Ministry of Education (Xiangtan University), Xiangtan Hunan 411105, China
  • Received:2012-08-13 Revised:2012-09-07 Online:2013-02-01 Published:2013-02-25
  • Contact: PENG Yinghui

摘要: 针对微脉瘤的灰度分布特性,提出一种新的微脉瘤检测算法。首先通过多尺度匹配滤波筛选出候选微脉瘤病变点,并作为种子点利用区域生长技术分割出病变区域;然后提取病变区域特征向量;最终采用Adaboost神经网络集成分类器检测真实的微脉瘤病变。在公开的ROC数据集测试表明,所提方法检测的平均正确率达到40.92%,优于以往的双环滤波和形态学方法。

关键词: 糖尿病性视网膜病变, 眼底图像, 神经网络集成, 匹配滤波, 微脉瘤检测

Abstract: According to the gray distribution characteristics of microaneurysms, a new microaneurysm detection algorithm was proposed. First, by multi-scale matched filtering, candidate microaneurysm lesions were picked out as seed points. And region growing technology was applied to segment the lesion areas. Then the features of the lesion areas were extracted. Finally the Adaboost neural network ensemble was designed to distinguish the real microaneurysm from all of the candidate lesions. The proposed method was tested on public ROC database. The experimental results show that the average detection accuracy is 40.92%, which is better than that of previous doublering filtering and morphological methods.

Key words: diabetic retinopathy, fundus photograph, neural network ensemble, match filtering, microaneurysm detection

中图分类号: