摘要
在眼底图像处理中,边缘检测技术是许多研究领域的重要预处理手段,如眼底图像特征提取和基于图像特征的图像配准方法等等。将希尔波特-黄变换(HHT,Hilbert-Huang Transform)方法应用于眼底图像边缘检测,提出了基于二维经验模式分解的图像边缘检测算法。实验结果表明,利用HHT方法的自适应多分辨特性,可以有效提取出眼底图像的细节信息,又不会检测出过多的虚假边缘。该方法为眼底图像的其它处理方法提供了有效的预处理手段。
Edge detection technology is an important pre-treatment means in many fields of fundus image processing,such as feature extraction or image registration based on the feature of images.Experiment proves the validity of above-mentioned methods.HHT is applied to edge detection of fundus image.A method of edge detection based on Bidimensional Empirieal Mode Decomposition（BEMD） is proposed.The experimental results shows that more image details could be retained and not too many false edges detected by the adaptive multi-resolution of HHT.The method provides an effective pre-treatment means for other treatment of fundus images.
出处
《通信技术》
2011年第11期71-72,75,共3页
Communications Technology
基金
福建省教育厅面上科技项目:EEG信号时空模型解耦与系统多元辨识(No.JA08006)
关键词
希尔波特-黄变换
经验模态分解
边缘检测
眼底图像
HHT（Hilbert-Huang Transform）
empirical mode decomposition
edge detection
fundus image
