Covid-19 diagnosis from x-ray images based on ensemble model with convolution neural network

COVID-19 is one of the many diseases that affects the lungs of the infected person. In 2020, the World Health Organization announced that COVID-19, began in China and propagated to many other countries, as a pandemic the disease effect the lungs and cues fever, cough, and fatigue. Moderate cases may have difficulty breathing or mild pneumonia (w.h.o) the identification of the pneumonia of patients infected by COVID-19 aids in preventing further spread and help us save lives of patients. The Convolution Neural Network proposed in this thesis provides early diagnosis of COVID-19 using X-ray images. X-ray imaging is cheap, fast and available in almost all hospitals. The Ensemble Model has a three convolution neural network that is trained on same data set, makes predictions on the same X-ray images and sums the result of the three models. Detailed comparisons have been done with the other state-of-the arts approaches. Proposed Ensemble Model provides up to 96% accuracy.