부산대학교 도서관

The article focuses on the research of image classification algorithms, namely the images indicate pathology of pneumonia caused by bacteria and viruses. The proposed method is based on using the VGG16, VGG19, DenseNet169 networks to extract data characteristics and train the model classification. The X-rays are classified including normal people, patients with viral pneumonia, and bacterial pneumonia. The provided source was medical data on chest X- ray images of patients who were manually classified by specialists. However, the accuracy of the classification is highly dependent on the number of images, the resolution of the images, and whether the X-ray image is correctly classified. In this study, the algorithms give relatively positive classification results with an accuracy of approximately 85%.