FEATURE EXTRACTION AND K-MEANS CLUSTERING APPROACH TO CLASSIFY THE COVID-19 LUNG CT-SCAN IMAGE

Feature extraction is the most important step in the classification process. Feature extraction is a method to obtain some statistical features about the image. The level of accuracy in the classification depends on the feature extraction. For detecting COVID-19, there are many features that can be used to classify them, including morphological feature extraction, first-order and second-order textures (GLCM). In this research, some features are used such as eccentricity, metric, mean, variance, skewness, contrast, correlation, energy, and homogeneity, which are then classified by the K-Means Method. The morphological feature data for cluster 1 is 98 data points and cluster 2 is 32 data points. The first-order texture feature data for cluster 1 is 88 data points, and cluster 2 is 42 data points. The last one uses GLCM data for cluster 1, and there are 75 data points, while cluster 2 has 55. From the calculation of accuracy, sensitivity, specificity, precision, and recall, the highest value is 50% for first-order texture extraction data, while the morphological feature extraction and GLCM data are 49.23% and 47.69%