M. A. Amer, M. Rihan, S. El-Agooz, Noha A. El-Hag, W. El-shafai, F. El-Samie, A. Khalaf, Ghada M. El-Banby, M. Abdelhamed
{"title":"Anomaly Detection System for Retinal Images Based on Area Classifier","authors":"M. A. Amer, M. Rihan, S. El-Agooz, Noha A. El-Hag, W. El-shafai, F. El-Samie, A. Khalaf, Ghada M. El-Banby, M. Abdelhamed","doi":"10.1109/ICEEM52022.2021.9480630","DOIUrl":null,"url":null,"abstract":"Diabetic Retinopathy (DR) is a disease of the eye for diabetics, and it can lead to a lack of vision if leaved untreated. The proposed approach in this paper is used to help for detecting and classifying the DR. It is applied to detect non-proliferative DR by identifying micro-aneurysms and hemorrhages. Firstly, the pre-processing step is applied. It consists of extracting the green channel, removing the optic disc (OD) and normalizing the background. Then, h-maxima transformation is performed. After that, threshold segmentation is applied to detect the hemorrhages and micro-aneurysms, accurately. Finally, an area classifier is used for the classification process to discriminate dark spot lesions.","PeriodicalId":352371,"journal":{"name":"2021 International Conference on Electronic Engineering (ICEEM)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 International Conference on Electronic Engineering (ICEEM)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICEEM52022.2021.9480630","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Diabetic Retinopathy (DR) is a disease of the eye for diabetics, and it can lead to a lack of vision if leaved untreated. The proposed approach in this paper is used to help for detecting and classifying the DR. It is applied to detect non-proliferative DR by identifying micro-aneurysms and hemorrhages. Firstly, the pre-processing step is applied. It consists of extracting the green channel, removing the optic disc (OD) and normalizing the background. Then, h-maxima transformation is performed. After that, threshold segmentation is applied to detect the hemorrhages and micro-aneurysms, accurately. Finally, an area classifier is used for the classification process to discriminate dark spot lesions.