视网膜血管图像分割及其应用研究

 摘要随着图像处理技术迅猛发展，人们能够更好治疗某些疾病以及愈后恢复情况的及时监控，从而使得诊疗水平得到极大的改善。图像计算机系统的建立，可以对眼底很多组织进行定量测量和检测，在正常和异常间做出明确鉴别，使系统的研究逐渐向标准化发展。本文围绕对视网膜血管分割的研究展开，首先对视网膜图像在多个方向上进行一系列的运算操作来进行滤波，这些运算主要包括形态学中膨胀、腐蚀、重建开运算和重建闭运算。通过这些预处理之后能够大大增强血管和背景的对比度，有效的降低了视网膜图像中的一些噪声，从而也提高了视网膜血管提取的精度。对于后期的视网膜配准和连接，主要是利用血管网络的中心线和交叉点为基础，采用血管网络相似性的配准方法，基本完成了图像的连接工作。在理论分析与研究的基础上，实现了视网膜血管的分割，基本完成了图像间的自动配准与连接。特别是分割取得了较好的效果，这为后续的图像配准和图像连接提供了良好的基础依据。关键词：图像处理技术；血管分割；图像配准；图像连接 AbstractWiththerapiddevelopmentofimageprocessingtechnology,peoplecanbetterprognosisforthetreatmentofcertaindiseasesandthetimelyrecoveryofcontrol,soitmaksthetreatmentbegreatlyimproved.Computerimagesystem,manyorganizationscanfundusquantitativemeasurementanddetection,betweennormalandabnormaltomakeaclearidentificationofthesystemgraduallytostandardization.Thisarticlefocusesonthesegmentationofretinalbloodvessels,thefirstimageontheretinaindifferentdirectionsonaseriesofoperationstocarryoutfilteringoperations,theseoperationsincludemorphologydilation,erosion,re-openedclosedoperationsandreconstructionoperations.Afterthesepretreatmentcanbegreatlyenhancedthecontrastofbloodvesselsandthebackground,effectivelyreducingtheretinalimageinsomeofthenoise,thusimprovingtheaccuracyofextractionofretinalbloodvessels.Lateregistrationforretinalandconnection,themainbloodvesselnetworkistousetheinterrectionofthecenterlineandthebasisofsimilaritywithvascularnetworkregistrationmethod,theimageoftheconnectionbasicallycompletedthework.Inthetheoreticalanalysisandresearchbasedontherealizationofthesegmentationofretinalbloodvessels,thebasiccompletionoftheautomaticregistrationbetweentheimageandlink.Inparticular,achievedgoodsegmentationresults,whichforthesubsequentimageregistrationandimagemosaic,providesagoodbasisfortheconnection.Keywords:imageprocessingtechnology;bloodvesseldetection;imageregistration;imagemosaic  目 录 摘要..................................................................................IABSTRACT...........................................................................II第一章绪论........................................................................11.1课题研究背景及意义..........................................................................................................................11.2国内外研究现状..................................................................................................................................21.3研究内容..............................................................................................................................................31.4本文的结构安排..................................................................................................................................4第二章研究基础.......................................................................52.1视网膜图像增强方法..........................................................................................................................52.1.1数学形态学理论基础..................................................................................................................52.1.2视网膜图像去噪..........................................................................................................................62.1.3视网膜血管增强..........................................................................................................................72.2图像分割..............................................................................................................................................72.2.1边缘检测................................................................................................................................................82.2.1.1Robert算子...................................................................................................................................82.2.1.2Prewitt算子和Sobel算子..............................................................................................................92.2.1.3Canny算子....................................................................................................................................102.2.2阈值分割..............................................................................................................................................102.2.2.1最佳阈值分割................................................................................................................................112.2.2.2类间方差阈值分割........................................................................................................................112.2.2.3最大熵阈值分割...........................................................................................................................122.3本章小结............................................................................................................................................13第三章视网膜血管分割...............................................................143.1视网膜图像特点分析........................................................................................................................143.2视网膜血管分割..........................................................................