Human Prakriti classification based on skin color using machine learning algorithms and image processing techniques

Abstract

In the field of Ayurveda, the term Prakriti represents an individual's intrinsic physiological and psychological constitution, which is traditionally categorized into three types: Vata, Pitta, and Kapha. Prakriti's precise identification is essential for providing personalized healthcare interventions. Nevertheless, conventional methods for assessing Prakriti are predominantly subjective and susceptible to variability. This study introduces an objective and automated methodology for classifying human Prakriti through the analysis of skin colour, utilizing image processing and machine learning techniques. High-resolution images of skin are subjected to pre-processing and feature extraction to generate colour histograms, texture metrics, and statistical colour parameters. A variety of machine learning classifiers, including K-Nearest Neighbours (KNN), Support Vector Machines (SVM) with Sigmoid, Linear, and Gaussian kernels, as well as Random Forest, are utilized for the classification task. Comparative evaluations indicate that the Random Forest and SVM with Gaussian kernel classifiers achieve superior accuracy, underscoring the efficacy of integrating image processing methodologies with advanced machine learning algorithms. The proposed system provides a standardized, efficient, and reliable approach to Prakriti classification, effectively merging traditional Ayurvedic concepts with contemporary technological innovations.