COVID-19 crisis has reshaped health globally and has increased the requirement for efficient testing strategies to catch infections promptly. This analysis looks into the application of AI strategies including CNNs in combination with texture-based feature extraction to detect COVID-19 automatically. CNN-Texture- Interaction Ensemble (CTIE) fuses CNNs and methods for texture feature extraction including the Gray-Level Co-occurrence Matrix (GLCM) to enrich feature characterization and advance classification efficiency. On the COVID-19 Radiography Database and COVIDx datasets the CTIE model obtained high accuracy rates of 94.3% and 93.5% respectively. High values of precision recall and F1 scores were indicated in metric calculations together with AUC scores of 95.2% and 94.5% for the datasets. The developed model exceeds the performance of traditional machine learning models including Support Vector Machines (SVM) and Random Forest alongside deep learning models that do not incorporate texture features. These results emphasize that combining CNNs with texture feature analysis improves COVID-19 detection accuracy.