ABSTRACT
The first incidence of Covid-19 was detected in Wuhan, China during December 2019. The SARS-CoV-2 virus causes Covid-19. Until recently, SARS-COV-2 was a dangerous virus that was spreading very fast.It has already threatened millions of lives and it continues. People with a covid ailment will also have mild - to - moderate symptoms and will be treated without the need for therapeutic intervention. However, some individuals will become extremely unwell and need medical help. RT-PCR approach has been shown to be effective in detecting coronavirus infection. The test results of RT-PCR may not be exactly true always. The alternative to the RT-PCR is detecting the virus through CT-Scan.It's based on radiography images and can help radiologists diagnose patients faster.Because this is a diagnosis based on images, the results will be more accurate and precise.The algorithm we have used in this paper is CNN. CNN transfer learning necessitates a limited amount of images to investigate. Also with this approach we can get higher accuracies around 90% to 98%. So our ultimate aim is to use these methods to detect if a person is infected with covid. This work have done the evaluation based on databases available and the results are compared. © 2022 IEEE.