Corona Virus Disease-19 is a crucial respiratory human-to-human communicable disease. Computer Tomography scan images are utilized for accurate detection. Even though enormous conventional methodologies are developed to detect the coronavirus disease-19, they failed to mitigate certain discrepancies including data scarcity issues, overfitting problems with increased computational complexity, and inability to obtain optimal convergence. Hence, an effective Ratel-optimized distributed contrastive N-pair loss-enabled Convolutional Neural network model (RDC2N) is proposed. The incorporation of the preprocessed via Fast Kernel Region Sharpening (FKRS) and segmentation with Ratel Adaptive hunt and Acquisition optimized multi-granular (RA2MG) approach stipulated the quality of the image for effective detection. Further, the feature extraction using the Discrete Invariant Geometrical Transform (DIGT) descriptor enhanced the coronavirus detection efficiency. In addition, the active tuning of the model and segmentation parameters using the Ratel Adaptive Hunt and Acquisition (RA2H) optimization attains optimal convergence. Moreover, the experimental results carried out by the SARS-COV-2 CT-Scan dataset achieved Matthew’s Correlation Coefficient of 0.92, Negative Predicted Value of 0.96, and Positive Predicted Value of 0.97 for 90% training percentage.