Multispectral wavebands selection for the detection of potential foreign materials in fresh-cut vegetables

Authors: TUNNY, SALMA - Chungnam National University; AMANAH, HANIM - Gadjah Mada University; FAQEERZADA, MOHAMMAD - Chungnam National University; WAKHOLI, COLLINS - Chungnam National University; BAEK, INSUCK - Orise Fellow; Kim, Moon; CHO, BYOUNGKWAN - Chungnam National University

Submitted to: Sensors
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/22/2022
Publication Date: 2/24/2022
Citation: Tunny, S., Amanah, H., Faqeerzada, M., Wakholi, C., Baek, I., Kim, M.S., Cho, B. 2022. Multispectral wavebands selection for the detection of potential foreign materials in fresh-cut vegetables. Sensors. 22:1775. https://doi.org/10.3390/s22051775.
DOI: https://doi.org/10.3390/s22051775

Interpretive Summary: Global demand for and consumption of fresh-cut vegetables have consistently increased and produce industries have sought technologies to help them maintain safe, high quality products. Nondestructive sensing technologies have been exploited as rapid means to assess safety and quality of fresh produce products. In this study, a near-infrared spectroscopic technique was investigated to detect and identify various types of foreign materials that can potentially be found in common fresh-cut vegetables. Multivariate analysis and modeling methods were applied to the spectral data to classify the fresh-cut vegetables and foreign materials. The modeling achieved 99% accuracy for the discrimination of the foreign materials from the fresh-cut vegetables under investigation. The research demonstrates the potential of the spectroscopic technique and provides insightful technical information that produce industries can use to detect foreign materials in fresh-cut vegetables.

Technical Abstract: Ensuring the quality and safety of fresh-cut vegetables is the greatest challenge for the produce industry and is equally important to both producers and to consumers. In this study, the possibility was investigated of using near-infrared spectral analysis as an effective technique to identify various types of FMs from seven common fresh-cut vegetables. All of the vegetables were cut into small pieces based on the size of the sample holder and the spectra for all the vegetable pieces and FMs were collected. To classify the FMs and vegetables, partial least square discriminant analysis (PLS-DA) and data driven soft independent modeling of class analogy (DD-SIMCA) model were developed, which had > 99% accuracy. The waveband selection algorithm, the variable importance of projection (VIP), showed that the VIP PLS-DA model achieved 99% accuracy for the discrimination of FMs from the various fresh-cut vegetables. The results indicate the high potential of the near-infrared spectroscopic technique to detect the FMs from fresh-cut vegetables for industrial application.