文章基本信息

标题：IMAGE CLASSIFICATIONS ON COVID 19 CXR IMAGES USING AUTO COLOR CORRELOGRAM FILTER
本地全文：下载
作者：S.M.Manimegalai ; T.Ramaprabha
期刊名称：Indian Journal of Computer Science and Engineering
印刷版ISSN：2231-3850
电子版ISSN：0976-5166
出版年度：2021
卷号：12
期号：5
页码：1288-1301
DOI：10.21817/indjcse/2021/v12i5/211205174
语种：English
出版社：Engg Journals Publications
摘要：In this current situation necessary to control the spread of virus by using computer-based analysis for the purpose of fast detection and reliable of corona virus disease (COVID-19) to reduce the issues on medical field. In image processing techniques, Chest X ray imaging are having more benefits like very cheap, portability, fast and easily can handle. This paper explores the effect of various popular image enhancement techniques. This research work finds that the Random Forest classifier is producing highest accuracy value precision value, recall value, Mathews correlation coefficient value, ROC value, PRC value and Kappa value which are 80.96% of accuracy level, 0.81 of precision value, 0.81 of recall value, 0.72 of Mathews Correlation value, 0.95 of Receiver Operating Characteristic value, 0.87 of Precision Recall value and 0.70 of kappa statistic value. It is having lowest Mean Absolute Error (MAE) value, Root Mean Squared Error (RMSE) value, Relative Absolute Error (RAE) value, Root Relative Squared Error (RRSE) value which is accordingly 0.14 of MAE value, 0.20 of RMSE value, 43.73% of RAE value and 64.40% of RRSE value. Finally this research work recommends that the Random Forest classifier is most recommend model by using Auto Color Correlogram Filter in image classifications.
关键词：Auto Color Correlogram Filter;BayesNet;Sequential Minimal Optimization;lazy classifier and Tree classifier