Under these circumstances, when we are still under the threat of various strains of coronavirus, RT-PCR is a long manual testing procedure that may generally result in a few hours to two days. This proposed system looks to create a model that uses a radiology based technique for detection. X-ray images are used for detecting the coronavirus. The aim of this study is to detect the patients whose x-rays are taken to check whether they are affected by COVID-19 or not. The dataset contains X-ray scans of people with Pneumonia, Normal and COVID-19 patients. This research work contains the implementation of VGG19, INCEPTION V3 and ResNet-50, the three algorithms, and a comparison of their results. For the image classification task, Convolution Neural Network (CNN) models take a crucial role. Evaluation of the model is performed using different metrics.