Application of computer vision on non-destructive detection of grape syrup adulteration

Grape is one of the most important garden products in the world because of its high waste some by-products like syrup are made from. The most important quality factor for grape syrup depends on its purity, which producers and consumers pay particular attention to. The grape syrup is one of the food products which are adulterated due to economic profit in the Iranian market. The development of simple, low cost, accurate and fast techniques to adulteration detection is essential in the food industry. In this study, the combination of image processing and nearest neighbor method (KNN) has been used for rapid and non - destructive adulteration detection of grape syrup. In this study, image processing combined with k-nearest neighbors are employed to fast and non-destructive adulteration detection of grape syrup. After image acquisition, the images are preprocessed and transformed into the RGB, HSI, and L*a*b* color spaces and finally textural statistical features are extracted from each image channels. In order to reduce the feature matrix dimension and increase the speed and accuracy of classification the principal component analysis (PCA) is applied. KNN is used for classifying image into four classes. Then statistical indexes such as accuracy, precision, sensitivity, specificity, and area under the curve are calculated to evaluate the model that the values of these indexes are obtained 96.25, 91.67, 91.19, 97.79, and 94.49 %, respectively, for test data. Therefore, the results show that this system has the ability to detect adulteration in pure grape syrup as a smart, fast, non-destructive and accurate method.