AUTOMATED IDENTIFICATION OF PALATAL RUGAE FROM DENTAL IMAGES USING MACHINE LEARNING CLASSIFIERS

Abstract

Palatal rugae are stable anatomical structures within the oral cavity that serve as reliable biometric markers in forensic odontology. This study presents a comprehensive computational pipeline for the automated identification of palatal rugae from occlusal images, combining image preprocessing, segmentation, feature extraction, and classification. We implement Support Vector Machines (SVM), Random Forests, and Convolutional Neural Networks (CNNs), demonstrating their performance on curated datasets. Experimental results indicate that CNN-based models outperform traditional classifiers when sufficient training data are available. Figures, tables, and code snippets illustrate the implementation pipeline.