Classification of Spices using Machine Learning Techniques

Machine learning (ML) has played a significant role in pattern recognition including fruits and vegetables classification. In this paper, comparative analysis of various ML techniques have been carried out for the identification of Spices. For the current work, ML techniques namely Naïve Bayes (NB), Decision Tree (DT), K-Nearest Neighbor (KNN), Random Forest (RF) and Support Vector Machine (SVM) have been undertaken. The main aim of the current study is to find out the most appropriate ML approach for Spices recognition. The experimental study has been performed on primary dataset of Spices. This dataset consists of 1000 images of five different Spices including clove, green cardamom, cinnamon, black pepper and curry leaf. The performance of the ML techniques have been analyzed on the basis of four parameters i.e. accuracy, precision, recall and f1-score. Out of five implemented ML models, best performance has been predicted by SVM approach with accuracy of 94.5%, precision of 95%, and recall of 94% with f1-score of 0.95..