{"title":"A new method for optic disc localization in retinal images","authors":"Manish Kumar Aggarwal, V. Khare","doi":"10.1109/IC3.2016.7880263","DOIUrl":null,"url":null,"abstract":"In this paper, a new method for localizing optic disc (OD) in retinal images has been proposed. Optic disc is an anatomical feature of retinal image which is considered as a first step for blood vessel segmentation and behaves as a land mark for retinal disease identification. Optic disc is a bright portion of retinal image and share similar properties with early pathology like exudates of Diabetic retinopathy disease. Thus, it is difficult to locate in retinal image due to the presence of similar pathology and non-uniform illumination in retinal image. In proposed approach pixel intensity is used as primary component and only bright pixel are considered for optic disc detection. Secondly, we have assumed that OD is a unique region in the retinal image having blood vessels converges on it. Therefore the optic disc can be localized by correlating every pixel value in the test image with a similar template having an OD at the centre. Finally a blood vessel map is extracted using large window sized median filter. The location of optic disc obtained is verified by the presence of blood vessel pixels in 10×10 window of the optic disc. The result of the proposed method is verified on two standard publicly available DRIVE [13] and DIARETDB1 [15] datasets comprises of 129 images, out of which optic disc in 126 retinal images is detected accurately.","PeriodicalId":294210,"journal":{"name":"2016 Ninth International Conference on Contemporary Computing (IC3)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2016-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"4","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2016 Ninth International Conference on Contemporary Computing (IC3)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IC3.2016.7880263","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 4
Abstract
In this paper, a new method for localizing optic disc (OD) in retinal images has been proposed. Optic disc is an anatomical feature of retinal image which is considered as a first step for blood vessel segmentation and behaves as a land mark for retinal disease identification. Optic disc is a bright portion of retinal image and share similar properties with early pathology like exudates of Diabetic retinopathy disease. Thus, it is difficult to locate in retinal image due to the presence of similar pathology and non-uniform illumination in retinal image. In proposed approach pixel intensity is used as primary component and only bright pixel are considered for optic disc detection. Secondly, we have assumed that OD is a unique region in the retinal image having blood vessels converges on it. Therefore the optic disc can be localized by correlating every pixel value in the test image with a similar template having an OD at the centre. Finally a blood vessel map is extracted using large window sized median filter. The location of optic disc obtained is verified by the presence of blood vessel pixels in 10×10 window of the optic disc. The result of the proposed method is verified on two standard publicly available DRIVE [13] and DIARETDB1 [15] datasets comprises of 129 images, out of which optic disc in 126 retinal images is detected accurately.