• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>International Journal of Innovative Computing Information and Control >BLOOD VESSEL SEGMENTATION IN RETINAL IMAGES BY MORPHOLOGICAL OPERATIONS AND BY A NOVEL PIXEL TRACKING ALGORITHM
【24h】

BLOOD VESSEL SEGMENTATION IN RETINAL IMAGES BY MORPHOLOGICAL OPERATIONS AND BY A NOVEL PIXEL TRACKING ALGORITHM

机译：形态操作和新型像素跟踪算法对视网膜图像中的血管进行分割

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Diabetic retinopathy is the most common diabetic eye disease and a leading cause of blindness. Diagnosis of diabetic retinopathy at an early stage can be done through the segmentation of the blood vessels of retina. Many studies have been carried out in the last decade in order to obtain accurate blood vessel segmentation in retinal fun-dus images including supervised and unsupervised methods. Supervised methods provide higher accuracies however they require more calculation time and their performance completely depend on the training data. In this work, an unsupervised segmentation method is proposed for blood vessel segmentation. A new vessel strengthening method and pixel tracking algorithm are presented. The pixel tracking algorithm filters background artifacts that have similar characteristics with small and thin vessels while segmenting vessels. The proposed segmentation method eliminates the dependency on training data and decreases the processing time of segmentation process while achieving high accuracy rates. Its performance is evaluated by using the DRIVE and STARE databases and achieved an average accuracy of 94.02% for the DRIVE database and 9543% for the STARE database. Our results demonstrated an average sensitivity of 71.74% and a specificity of 97.28% for the DRIVE database while it is 71% and 97.4% for the STARE database, respectively. The calculated sensitivity and specificity values for both databases also state that the proposed segmentation method is reliable.
机译：糖尿病性视网膜病是最常见的糖尿病眼病，也是失明的主要原因。糖尿病视网膜病变的早期诊断可以通过视网膜血管的分割来完成。在过去的十年中已经进行了许多研究，以便在视网膜有趣的图像中获得精确的血管分割，包括有监督和无监督的方法。有监督的方法具有较高的准确性，但是它们需要更多的计算时间，并且其性能完全取决于训练数据。在这项工作中，提出了一种无监督的血管分割方法。提出了一种新的血管加固方法和像素跟踪算法。像素跟踪算法在分割血管时过滤具有细小血管相似特征的背景伪像。提出的分割方法消除了对训练数据的依赖，并减少了分割过程的处理时间，同时实现了较高的准确率。通过使用DRIVE和STARE数据库评估其性能，对DRIVE数据库的平均准确性为94.02％，对于STARE数据库的平均准确性为9543％。我们的结果表明，DRIVE数据库的平均灵敏度为71.74％，特异性为97.28％，而STARE数据库的平均灵敏度分别为71％和97.4％。计算出的两个数据库的敏感性和特异性值也表明，提出的分割方法是可靠的。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号