%0 Journal Article
%T 基于LPQ特征的视网膜OCT图像分类算法<br>Algorithm for Classification of the Retinal OCT Images with LPQ Features
%A 任岚
%A 穆国旺
%J Computer Science and Application
%P 112-117
%@ 2161-881X
%D 2020
%I Hans Publishing
%R 10.12677/CSA.2020.101012
%X <div style=\"text-align:justify;\">
	<span style=\"font-size:14px;\">为缓解我国眼科疾病患者多、医生少、医疗压力巨大的国情，提出一种基于局部相位量化(local phase quantization, LPQ)特征的视网膜OCT图像分类算法。首先对图像进行预处理，主要包括对感兴趣区域探测阶段、拟合阶段和切割阶段；其次提取切割后图像的LPQ特征；然后利用PCA方法对其降维；最后，利用SVM进行分类。在Duke视网膜数据集上对算法进行了验证，并和现有文献中提到的LBP特征、Gabor特征及SIFT特征进行了对比研究。实验结果表明，利用LPQ特征可以得到相对更好的分类结果。<br />
In order to alleviate the situation of more ophthalmology patients, few doctors and huge medical pressure in China, an algorithm for classification of retinal OCT images with LPQ features is proposed. Firstly, the acquired OCT images are preprocessed in three steps including the perceiving phase, the fitting phase and the cutting phase; Secondly, LPQ features are extracted; Then PCA method is used to reduce the dimensionality; Finally, the SVM is employed for classification of images. The algorithm is verified on Duke retinal data set, and is compared with those methods which use LBP feature, Gabor feature or SIFT feature. Experimental results show that the LPQ feature can obtain relatively better classification results.</span> 
</div>
%K 计算机辅助诊断，视网膜OCT图像，局部相位量化，主成分分析，支持向量机<br>Computer Aided Diagnosis
%K Retinal OCT Image
%K Local Phase Quantization
%K Principal Component Analysis
%K Support Vector Machine
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=34009