Spices are usually the most widely used ingredients in the food industry, which are sold in bulk and powdered form in the market, and the possibility of adulterating them is very high. Therefore, the main reason for conducting this piece of research is the demand for new and accurate methods that measure the authenticity and quality of spices (such as turmeric powder) through electronic nose and electronic tongue systems. Unmarketability of 3rd grade chickpeas and their very low price has become an excuse for some profiteers to mix them with turmeric powder and sell them in the market. In this research, samples of turmeric powder were prepared in 13 classes and electronic nose and electronic tongue systems were used to evaluate their quality and detect fraud. Finally, the results obtained from the output of these systems were evaluated and classified using machine learning algorithms (such as basic CNN, improved CNN, PCA, MLP, Fuzzy, SVM, KNN, GBT and EDT). The results showed that the electronic nose and electronic tongue systems in combination with the improved CNN are able to classify the samples of turmeric powder and fraud with 94.87% and 95.38% accuracy, respectively. The PCA method in the electronic nose system for evaluating the samples of turmeric powder and fraud obtained 96% and in the electronic tongue system 97% of the variance between the studied samples.
Jahanbakhshi et al. (Mon,) studied this question.