Coronavirus disease (COVID-19) appeared in the last quarter of 2019. A Coronavirus wildfire around the world, where the infection and death rates uprise dramatically every day. In this paper, an enhanced convolutional neural network based on transfer learning was proposed to detect patients infected with COVID-19 using chest X-ray radiographs. The proposed model helps radiologists to diagnose COVID-19 disease automatically with high accuracy. The proposed method is introduced to afford precise identification of infected persons. Referring to the performance results acquired, the used pre-trained ResNet50 paradigm achieves the peak performance with 99.2% accuracy outperforming all compared methods.