The presence of hidden fungal disease inside pomegranate fruit has reduced the price in the trade of the pomegranate. Alternaria spp. is a widespread fungal disease threatening pomegranate quality. The present study aimed to examine the efficiency of the Electronic nose (E-nose) system as a fast, nondestructive, and low-cost method in diagnosis the amount of Alternaria fungi of the pomegranate. Sixty samples were classified to 0, 25, 50, 75, and 100% amount of Alternaria spp. Linear Discriminant Analysis (LDA), Back Propagation Neural Network (BPNN) and Support Vector Machine (SVM) methods were applied and compared as linear and non-linear analysis methods for detection. The results showed that the LDA method successfully detected healthy and infected samples with 100% accuracy, only by using two Metal Oxide Semiconductor (MOS) sensors. As a prediction method, BPNN showed higher accuracy of 100% in the detection of 0, 25, 50, 75, and 100% infected pomegranates. The results indicated that the E-nose technique is a reliable instrument to detect the quality of the pomegranate with high precision.Abbreviations: BPNN: back propagation neural network; E-nose: electronic nose; LDA: linear discriminant analysis; MOS: metal oxide semi-conductor; PCA: principal component analysis; SVM: support vector machine ARTICLE HISTORY
Chocolate is one of the most consumed products in the world. Chocolate is a mixture of roasted cocoa beans and ultrafine particles of sugar (or without containing cocoa butter). The mixture is supplied with different percentages of cocoa in the market. In this research, the ability of an olfactory system is evaluated as a fast, nondestructive, and low‐cost method to detect bitter chocolates with different cocoa percentages. For this purpose, bitter chocolates with three different percentages of cocoa (78, 85, and 96%) were selected for evaluation. Separation rate of electronic nose (E‐nose) dataset obtained 99%, by using principal components analysis method. To assess the classification accuracy of the bitter chocolate samples, linear discriminant analysis, and support vector machine applied. The results showed that both methods were able to classify bitter chocolates of different percentages of cocoa with the highest accuracy of 100%. This research demonstrates the potential of using the olfactory system as a fast, nondestructive, low‐cost, and reliable method for distinguishing and categorizing different cocoa percentages in bitter chocolates with high precision.
Practical applications
Most researches were focused on a specific percentage of cocoa, while there are few reports on the detection of different percentages of cocoa of the same type. In our study, the linear discriminant analysis and support vector machine methods were employed to the distinguishing of various percentages of cocoa of same type (bitter chocolate), and it was observed that both methods achieved the better classification performance, which indicated the feasibility of distinguishing different percentages by E‐nose. It is well‐known that the detection of different cocoa percentages from the same type would take a long time and require developed instrumental laborites with experts, thus it is of great significance to establish a rapid nondestructive detection method for the related study in the field of the cocoa industry.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.