V. S. S. P. R. Gottumukkala, N. Kumaran, V. Sekhar
{"title":"Skin Lesion Segmentation Using SCU-Net with FNLM Preprocessing","authors":"V. S. S. P. R. Gottumukkala, N. Kumaran, V. Sekhar","doi":"10.1109/ICDSIS55133.2022.9915935","DOIUrl":null,"url":null,"abstract":"Recently, people are suffering with variety of skin cancers due to radiation problems, atmospheric effects and change in environmental conditions. So, early detection of skin cancers can save the millions of people. The conventional image processing methods were failed to localize disease effected region accurately, which caused improper prediction of skin cancer types. Therefore, this article is implemented the preprocessing-based skin lesion segmentation network (SLS-Net). Initially, fast nonlocal mean (FNLM) filter is applied to remove the different types of noises from skin lesions, which also enhances the skin lesion image. Further, skip connection-based U-Net (SCU-Net) model is applied for accurate segmentation of skin lesions. The simulations performed on ISIC-2019 and PH2 datasets discloses the superiority of proposed SLS-Net in terms of precision, recall, sensitivity, specificity, and f1-score as compared to conventional segmentation models.","PeriodicalId":178360,"journal":{"name":"2022 IEEE International Conference on Data Science and Information System (ICDSIS)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE International Conference on Data Science and Information System (ICDSIS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICDSIS55133.2022.9915935","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
Recently, people are suffering with variety of skin cancers due to radiation problems, atmospheric effects and change in environmental conditions. So, early detection of skin cancers can save the millions of people. The conventional image processing methods were failed to localize disease effected region accurately, which caused improper prediction of skin cancer types. Therefore, this article is implemented the preprocessing-based skin lesion segmentation network (SLS-Net). Initially, fast nonlocal mean (FNLM) filter is applied to remove the different types of noises from skin lesions, which also enhances the skin lesion image. Further, skip connection-based U-Net (SCU-Net) model is applied for accurate segmentation of skin lesions. The simulations performed on ISIC-2019 and PH2 datasets discloses the superiority of proposed SLS-Net in terms of precision, recall, sensitivity, specificity, and f1-score as compared to conventional segmentation models.