Background & Purpose: Due to the lack of definitive and effective treatment and the unknown side effects of Covid-19 disease, disruption of the transmission chain and the prevalence of this disease is of great importance. Providing intelligent and automated image analysis solutions is an effective step in controlling this disease that can reduce the workload in the field of health.

Methodology: This study introduced and examined the specific features of coronavirus in lung CT scan images, and processing and intelligent diagnostic methods (using machine learning method) to identify Covid-19 patients.

Findings: In this study, after receiving CT scan images of the lungs, first pre-processing is performed and then with the help of GLCM and HOG algorithms, feature extraction is performed from the images. Due to the large number of image extraction properties, the classification operation is not possible, so with the help of algorithm Chi-Square reduces the number of properties extracted and then uses the XGBoost algorithm to classify the extracted properties and distinguish between sick and healthy people.

Conclusion: Using a data set, 2484 CT scan images of healthy and sick people and using evaluation parameters, this study has achieved the results: 92.2% accuracy, 92.4% sensitivity and 92.3% F score.