Automated caries detection based on Radon transformation and DCT

Dental Caries is caused by the infection in the calcified tissue of the teeth. This paper proposes an automated caries detection system based on Radon Transformation (RT) and Discrete Cosine Transformation (DCT). The Radon Transformation (RT) is performed on these X-Ray images for each degree to capture the low frequency details. Then 2-D DCT is applied to RT images to obtain the frequency features (DCT coefficients). These features are further converted to 1-D coefficient vector in Zigzag fashion which is subjected to Principal Component Analysis (PCA) for feature extraction. Finally, minimum number of features are fused using Decision Tree (DT), k-Nearest Neighbour (k-NN), Random Forest, Naive Bayes, Sequential Minimum Optimization (SMO), Radial Basis Function (RBF), Decision Stumps, AdaBoost classifiers to get highest classification performance.