Rapid and nondestructive detection of freshness quality of postharvest spinaches based on machine vision and electronic nose

• Published in: Journal of food safety Vol. 39; no. 6
• Main Authors: Huang, Xingyi, Yu, Shanshan, Xu, Haixia, Aheto, Joshua H., Bonah, Ernest, Ma, Mei, Wu, Mengzi, Zhang, Xiaorui
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.12.2019; Blackwell Publishers Inc
• Subjects: Accuracy; Analytical methods; Artificial neural networks; Classification; Cold storage; Data integration; Decay; Economic analysis; Economic impact; Electronic noses; Food safety; Freshness; Machine vision; Minerals; Model accuracy; Multisensor fusion; Neural networks; Nondestructive testing; Nutrients; Odor; Odors; Organic chemistry; Physicochemical analysis; Refrigeration; Sensory evaluation; Spinach; Support vector machines; Vision systems; Vitamins
• ISSN: 0149-6085; 1745-4565
• DOI: 10.1111/jfs.12708

Artificial sensory evaluation and physicochemical analysis tend to be tedious and time consuming. Therefore, this study employs machine vision, electronic nose (E‐nose), and multisensory data fusion to rapidly and nondestructively detect spinach freshness during storage. Spinach freshness during refrigeration was classified into four grades by 10 trained panelists. Machine vision and E‐nose were then used to obtain the image and odor information of the samples, respectively. K‐nearest neighbor (KNN), support vector machine (SVM), and back‐propagation artificial neural network (BPNN) were applied to predict spinach freshness. Result from the BPNN model based on machine vision achieved the same result as the KNN model with a classification accuracy of 85.42%. The BPNN model based on E‐nose achieved a better result than the SVM model with classification accuracies of 81.25% and 75.00%, respectively. Furthermore, the BPNN model based on multisensory data fusion greatly improved the detection accuracy of spinach freshness with a classification accuracy of 93.75%. Multisensory data fusion approaches based on machine vision and E‐nose are capable of intelligently detecting the freshness of postharvest spinach during storage. The main advantage is to allow a rapid and nondestructive detection of spinach freshness without using any chemical pretreatments. Practical Applications Fresh spinach is an important source of minerals, vitamins, carotenes, and other nutrients. Timely and accurately knowing the grade of spinach freshness before decay is of great interest in ensuring edible quality and reducing economic losses. Traditional analytical methods of spinach freshness tend to be tedious and time consuming. Novel analytical techniques such as machine vision and electronic nose (E‐nose) showed good feasibility in detecting spinach freshness during refrigerated storage. The detection accuracy could be highly improved based on the multisensory data fusion of machine vision and E‐nose. The results of this study confirmed that machine vision and E‐nose could constitute a rapid and nondestructive method to detect spinach freshness during storage instead of tedious and time‐consuming methods. The method of combining machine vision and E‐nose could be applied in the fruit and vegetable industry in the future.