Coronavirus (CV) disease (COVID-19), caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), has emerged as one of the most significant global health challenges. Since its outbreak, COVID-19 has resulted in over 5 million deaths worldwide and continues to pose a serious public health threat due to the virus’s rapid mutation and emergence of highly transmissible variants. Early detection of COVID-19 is crucial to reducing mortality rates and controlling its spread. However, traditional diagnostic techniques face limitations, including limited availability, high cost, and time-consuming procedures. Medical imaging, in combination with machine learning (ML) and deep learning (DL) techniques, has shown promise in diagnosing respiratory disorders, including COVID-19. While conventional ML methods rely on hand-crafted features, the performance of these classifiers is highly dependent on the quality and relevance of the extracted features. Here, the COVID-19 classification utilizing CT images is done using the Gradient Descent based DenseNet (GD-DenseNet). The phases conducted to perform COVID-19 classification are preprocessing, feature extraction (FE) and classification. At first, considered CT scan image is preprocessed employing Adaptive Bilateral Filter (ABF). Then, the extraction of features like Local Directional Pattern (LDP), Gray-level co-occurrence matrix (GLCM) and Convolutional Neural Network (CNN) features are done. Finally, the DenseNet is used for COVID-19 classification, where the training is done using the Gradient Descent (GD) algorithm. Moreover, GD-DenseNet achieved maximal values of accuracy of about 88.51%, true negative rate (TNR) about 90.43% and true positive rate (TPR) about 86.87%.