KDD ON CXR Images of COVID 19 by JPEG Coefficient Filter Technique

Abstract

Prevent the spread of virus utilizing computer-based analysis for the goal of fast and reliable identification of corona virus disease (COVID-19). Chest X-ray imaging has many advantages in image processing methods, such as low cost, portability, speed, and ease of use. The impact of many image enhancement techniques are investigated in this research. The impact of many image enhancement techniques are investigated in this research. The impact of many common picture enhancement techniques is investigated in this research. The Random Forest (RF) classifier produces the highest accuracy value, positive predictive value, true positive rate, Area under the receiver operating characteristic curve value, Area under the precision recall curve value which are 84.03% accuracy level, 0.86 of positive predictive value, 0.84 of true positive rate value level, 0.96 of area under the receiver operating characteristic curve value, and 0.91 of area under the precision recall curve value This research work finds that the Random Forest learning model is most recommended model with JPEG Coefficient Filtering technique for CXR image classifications of COVID -19 Image dataset.