Non-destructive detection pilot study of vegetable organic residues using VNIR hyperspectral imaging and deep learning techniques

Authors: SEO, YOUNGWOOK - Korean Rural Development Administration; KIM, GIYOUNG - Korean Rural Development Administration; LIM, JONGGUK - Korean Rural Development Administration; LEE, AHEONG - Korean Rural Development Administration; KIM, BALGEUM - Korean Rural Development Administration; JANG, JAEKYUNG - Korean Rural Development Administration; MO, CHANYEUN - Kangwon National University; Kim, Moon

Submitted to: Sensors
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/19/2021
Publication Date: 4/21/2021
Citation: Seo, Y., Kim, G., Lim, J., Lee, A., Kim, B., Jang, J., Mo, C., Kim, M.S. 2021. Non-destructive detection pilot study of vegetable organic residues using VNIR hyperspectral imaging and deep learning techniques. Sensors. 21(9):2899. https://doi.org/10.3390/s21092899.
DOI: https://doi.org/10.3390/s21092899

Interpretive Summary: Contamination of processing equipment surfaces in produce processing facilities is a food safety concern. Various non-destructive imaging technologies have been investigated for detecting contaminants on foods and food-contact surfaces. In this study, a visible and near-infrared hyperspectral imaging technique was used to detect and classify a range of diluted potato and spinach residues on food processing equipment surfaces. For image processing and analysis, six methods including a convolutional neural network were evaluated. The results show that the range of diluted sample residues can be detected and classified with over 90% accuracy. The sensing technique presented in this study could benefit the food processing industry by providing a rapid means to detect contaminants on food-contact surfaces, and thereby reduce the food safety risks associated with cross contamination of food products.

Technical Abstract: Contamination is a critical issue that affects food consumption adversely. Therefore, efficient detection and classification of food contaminants are essential to ensure food safety. This study applied a visible and near-infrared (VNIR) hyperspectral imaging technique to detect and classify organic residues on the metallic surfaces of food processing machinery. The experimental analysis was performed by diluting both potato and spinach juices to six different concentration levels using distilled water. The 3D hypercube data were acquired in the range of 400–1000 nm using a line-scan VNIR hyperspectral imaging system. Each diluted residue in the spectral domain was detected and classified using six classification methods, including a 1D convolutional neural network (CNN-1D) and five pre-processing methods. Among them, CNN-1D exhibited the highest classification accuracy, with a 0.99 and 0.98 calibration result and a 0.94 validation result for both spinach and potato residues. Therefore, in comparison with the validation accuracy of the support vector machine classifier (0.9 and 0.92 for spinach and potato, respectively), the CNN-1D technique demonstrated improved performance. Hence, the VNIR hyperspectral imaging technique with deep learning can potentially afford rapid and non-destructive detection and classification of organic residues in food facilities.