ABSTRACT
In the course of an epidemic season of COVID-19, the CT image framework is expedient for detecting the COVID-19 suspected patients. The process of identification of CT images because of COVID-19 is a two-step process. Firstly, we apply to pre-process on input COVID-19 CT images. Secondly, we evaluated individual deep CNN models on a pre-processed image is COVID-19 positive or negative then compared the execution of the individual deep CNN models. Finally, we identified that the InceptionV3 model is best among all the deep CNN models, including InceptionV3, MobileNet, MobileNetV2, ResNet50, ResNet50V2, and Xception. It can be witnessed that the InceptionV3 prototype has provided high accuracy among all the prototypes. InceptionV3 has provided accuracy, f1-score, and AUC as 0.864864, 0.873417, and 0.86425, respectively. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.