ABSTRACT
Covid illness (COVID-19), with a beginning stage in China. As per report collected on 30 Jun 2022, total cases are 4.35Cr +18,819 also 5.25L+39 cases of deaths across several states of India are reported. it means quite a bit to complete a modified area structure as an expedient elective finding decision to hinder COVID-19 spreading among people. In this research, a novel method is proposed in which a efficient hybrid technique is developed to recognize Coronavirus utilizing half breed Chest and Lung's network based models. CNN and ResNet50 layers have been implemented in proposed HNN model for the acknowledgment of Covid pneumonia tainted patient using chest and lung X-shaft and lung radiographs. We have executed multi class portrayals with four classes (COVID-19, common (strong), viral pneumonia) by using 5-overlay cross endorsement. Considering the show results gained, it has seen that the pre-arranged ResNet50 model gives the most important portrayal execution (99.7% precision for Dataset). © 2022 IEEE.