When a person infected by COVID-19, the first damaged organ is a respiratory system and lungs. Early diagnosis of lung state increases the patient's chance to survive.we prepared our data set on 3 Subsets : train &; validation &; test. the data extracted from chest-xray images from Kaggle. we reorganised the data into 3 Clacess : Healthy that contain x-ray images for Normal person who's they not have any problem or Infection by virus or had bacteria. the second Class Infected-virus that contain x-ray images for patients who was infected by virus. the third Class is Bacteria consern patients import bacteria. if the prediction is healthy it meens that a person is normal and no contaminated . but if the prediction is Infected by virus or Bacteria it meens that a person is infected and must enter in a Quarantine. this dectection helps for Containment.