The first COVID-19 instances were reported in December 2019 by Wuhan City, Hubei Province, China (Xu et al.). Since the COVID-19 worldwide pandemic affects health care and lifestyles all over the world, early diagnosis is crucial to stopping the spread of cases and lowering mortality. Backward transcription The Polymerase Chain Reaction (RT PCR) is the gold standard in diagnostic testing; yet, due to its exorbitant costs and lengthy turnaround times, alternate quick and readily accessible diagnostic approaches are necessary. Current deep learning models (CNN) are used in this paper’s method to analyze the images and classify them as positive or negative for COVID-19. This method was motivated by recent research that connects the presence of COVID-19 to findings in Chest X-ray images. The suggested system includes a preprocessing stage that includes lung segmentation, eliminating the surrounds that do not provide pertinent information for the job and may create biased findings; Performance evaluation of the classification model using a confusion matrix and performance measures comes after this initial phase in which it was trained using a transfer learning technique. The most precise models had a 97% accuracy in detecting COVID-19. The improvement from the other models is then demonstrated by comparing our model with the current model for COVID-19 detection.