Large number of people around world were affected by global outbreak of coronavirus infection (COVID-19) and resulted in significant death worldwide. In this context, artificial intelligence has emerged as effective tool for early assessment and further analysis. In many cases, chest X-rays had played important and key role for early detection of COVID where image processing with machine learning can play very important role. The primary aim of our research is to find any computer-aided diagnostic method using X-ray images, deep learning method for COVID-19 detection. We used Convolution Neural Networks (CNN) ResNet50 for training of dataset and further applied top performing machine learning algorithms. Our algorithm was further evaluated on an independent test dataset. The chosen algorithms, combined with support vector machine (SVM) and machine learning features, exhibited very good suitability for distinguishing COVID, which boast and increase severity detection sensitivity and specificity in range 96.85–98.02%, while also maintaining robust performance across varying severity levels. Proposed method eliminates the need for manual interventions such as sample collection, enabling seamless integration with standard X-ray reporting model with help of advance machine learning. This integration offers an efficient, effective, and early detection solution for COVID-19 to address global healthcare challenge.