Development and Evaluation of an Intelligent Machine Vision System for Detecting Chicken Gizzard Adulteration in Minced Red Meat

Document Type : Research Paper

Authors

1 Mechanical Engineering of Biosystems Department, Faculty of Agriculture, Shahrekord University, Shahrekord, I.R.Iran

2 Biosystems Engineering Department,, Sari Agricultural Sciences and Natural Resources University, Sari, Iran

3 Mechanical Engineering of Biosystems Department, Faculty of Agriculture, Shahrekord University, Shahrekord,

4 Mechanical Engineering of Biosystems Department, Faculty of Agriculture, Shahrekord University, Shahrekord, I.R. Iran.

10.22059/ijbse.2025.398182.665600

Abstract

In this study, an intelligent machine-vision system was developed and evaluated for detecting chicken gizzard adulteration in ground red mutton-veal meat, using digital images captured by a mobile phone. To this end, standard samples of minced red meat (55% mutton and 45% beef) with varying proportions of chicken gizzard (0 to 100%) were prepared, and images were captured in a laboratory environment both directly from the sample surface and through plastic wrap packaging. The color features (RGB) of the images were extracted using the MATLAB image processing toolbox, and modeling was performed employing statistical and machine learning methods, including Principal Component Analysis (PCA), Partial Least Squares Regression (PLSR), and Multilayer Perceptron (MLP) neural networks. The best performance of the linear PLSR model for estimating the adulteration percentage yielded an R²V=0.7 and RMSEV=16.51 under conditions without plastic wrap, whereas the nonlinear MLP model achieved an R²V=0.97 and RMSEV=6.673. These results were lower (0.9 and 12.54) for the data acquisition through plastic wrap due to the light reflections caused by the covering. Furthermore, the MLP classifier achieved classification accuracies of 85%, 96.4%, 92.6%, 73.7%, 76.2%, and 96.7% for adulteration levels 0-10%, 10–20%, 20–30%, 30–40%, 40–50%, and above 50%, respectively. The average precision, sensitivity, and F1 score for the developed model were obtained as 0.975, 0.974, and 0.975, respectively. The results showed that this non-destructive method is both fast and reliable for identifying gizzard fraud in ground meat, and can serve as a basis for developing meat quality control systems.

Keywords

Main Subjects


 

Introduction

Food adulteration is a significant issue that impacts food safety, consumer trust, and regulatory compliance globally. One common form of adulteration in meat products involves the unauthorized addition of cheaper or lower-quality ingredients to higher-value meats. Specifically, the adulteration of chicken gizzard—a low-cost offal—into minced red meat, such as beef and mutton, poses both economic and health concerns. Traditional methods for detecting such adulteration often involve destructive, time-consuming, and costly laboratory analyses, which limit their applicability for rapid quality control.

This study aims to develop and evaluate an intelligent machine vision system capable of detecting chicken gizzard adulteration in minced red meat through non-destructive and rapid analysis. The core objective is to assess the feasibility of using digital color images captured by a mobile phone to identify and quantify the presence of chicken gizzard in mixtures of beef and mutton meat. By leveraging advances in image processing and machine learning, this approach offers a cost-effective and efficient solution for detecting food fraud.

Materials and Methods

To simulate real-world adulteration scenarios, standard minced red meat samples were prepared with a fixed base composition of 55% mutton and 45% beef. Chicken gizzard was added to these base samples in varying proportions, ranging from 0% (pure red meat) to 100% (pure chicken gizzard), creating a comprehensive gradient of adulteration levels for analysis.

Image acquisition was performed in a typical laboratory environment to evaluate the system’s robustness under less controlled circumstances. Each sample was imaged in two ways: directly from the exposed meat surface and through transparent plastic wrap packaging, simulating commercial retail conditions.

The color features of the captured images were extracted in the RGB (Red, Green, Blue) color space using MATLAB’s image processing toolbox. These color components served as the primary input variables for subsequent modeling.

Data analysis involved a combination of statistical and machine learning techniques. Principal Component Analysis (PCA) was first applied to visually discriminate the samples and identify the most significant color features related to adulteration. Partial Least Squares Regression (PLSR), a linear modeling technique, was then used to predict the percentage of chicken gizzard adulteration based on the extracted RGB features. To capture nonlinear relationships and enhance prediction accuracy, a Multilayer Perceptron (MLP) neural network was also developed and trained on the dataset.

Results

The analysis revealed a clear trend in color changes corresponding to increasing levels of chicken gizzard adulteration. This shift in color composition reflects the intrinsic differences in tissue pigmentation between red meat and chicken gizzard, providing a reliable basis for discrimination.

Quantitative evaluation of the models showed that the linear PLSR model achieved a coefficient of determination (R²) of 0.7 and a root mean square error (RMSE) of 16.51 when predicting adulteration levels in images captured without plastic wrap. While this indicates moderate accuracy, the nonlinear MLP model significantly outperformed PLSR, achieving an R² of 0.97 and RMSE of 6.67 under the same conditions. This demonstrates the superior capability of nonlinear modeling in capturing complex relationships between color features and adulteration levels. These results were lower (0.9 and 12.54) for the data acquisition through plastic wrap due to the light reflections caused by the covering. Furthermore, the MLP model was implemented as a classifier to categorize samples into discrete adulteration intervals: 0–10%, 10–20%, 20–30%, 30–40%, 40–50%, and above 50%. The classification accuracies achieved were 85%, 96.4%, 92.6%, 73.7%, 76.2%, and 96.7%, respectively. The average precision, sensitivity, and F1 score for the developed model were obtained as 0.975, 0.974, and 0.975, respectively. These results indicate that the system can reliably detect levels of adulteration ranging from low to high.

Conclusion

This study demonstrates that color image analysis combined with advanced machine learning techniques offers a non-destructive, rapid, and reliable method for detecting chicken gizzard adulteration in minced red meat. The developed machine vision system effectively distinguishes adulterated samples and accurately estimates the level of adulteration, particularly when employing nonlinear models such as MLP neural networks. The ability to perform such analysis through images captured even under less controlled lighting conditions and through plastic wrap packaging highlights the practical applicability of the system in real-world food quality control environments. This approach can serve as a foundation for developing automated, software-based quality assurance tools in the food industry, enhancing the detection of food fraud and protecting consumer interests. Future work may focus on expanding the system to detect other types of adulterants, integrating hyperspectral imaging techniques for improved sensitivity, and developing user-friendly interfaces to facilitate adoption by food producers and regulatory agencies.

Authorship Contribution

Mobin Rezazadeh: Writing – original draft, Methodology, Data curation, Software, Formal analysis.

Sajad Kiani: Writing – review and editing, Formal analysis, Investigation, Data curation, Validation.

Mahdi Ghasemi Varnamkhasti: Review and edit, conceptualize, supervise, and manage project administration.

Zahra Izadi: Advisor, methodology, and resources.

Data Availability

Data available on request from the authors. All the data used in this original research are presented throughout the text and in the form of Tables and Figures.

Ethical considerations

The authors avoided data fabrication, falsification, plagiarism, and misconduct.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflict of interest

The author declares no conflict of interest.

Azari, N.Z, Mahmoudian, F., & Shahsuni Mojard, L. (2021). Meat Products Industries: Studying the Nutritional Value of Red Meat. Fourth International Conference on Interdisciplinary Studies in Food Industries and Nutritional Sciences of Iran, Tehran, https://civilica.com/doc/1262624
Ballin, N.Z. (2010). Authentication of meat and meat products. Meat science, 86 (3), 577-587.‏
Chen, H., Tan, C., Lin, Z. (2025). Identification of beef adulteration based on near-infrared spectroscopy and an ensemble of radical basis function network. Journal of Food Composition and Analysis, 143, 107633.
Chenyu, L., Xinlian Z., Tanya T.N, Jinyuan L., Tsungchin W., Ellen L., Xin M.T. (2022). Partial least squares regression and principal component analysis: similarity and differences between two popular variable reduction approaches. General Psychiatry, 35:e100662. https://doi.org/10.1136/gpsych-2021-100662
Dean, T., Brendan, M., Gerard, D. (2005). Using Unlabelled Data to Update Classification Rules with Applications in Food Authenticity Studies. Journal of the Royal Statistical Society Series C: Applied Statistics, 55 (1), 1–14. https://doi.org/10.1111/j.1467-9876.2005.00526.x
Dehnavi Dehghan, M. A., Kohzadi, N., and Khalilian, S. (2005). Investigating structural change in meat demand through nonparametric revealed preferences test. Fifth Agricultural Economics Conference, University of Sistan and Baluchestan.
Dosti Irani, O., & Golzarian, M.R. (2013). Design and evaluation of image processing algorithm for estimating fat content of red meat. 8th National Congress of Agricultural Machinery Engineering (Biosystems) and Mechanization of Iran, Mashhad, https://civilica.com/doc/284547
Edris, M., Ghasemi-Varnamkhasti, M., Kiani, S., Yazdanpanah, H., & Izadi, Z. (2024a). Identifying the authenticity and geographical origin of rice by analyzing hyperspectral images using unsupervised clustering algorithms. Journal of Food Composition and Analysis, 125, 105737.
Edris, M., Ghasemi-Varnamkhasti, M., Kiani, S., Yazdanpanah, H., & Izadi, Z. (2024b). Application of fuzzy clustering algorithm and hyperspectral images for rice authentication. Iranian Journal of Biosystems Engineering, 55(2), 79-92.
Gewers, F.L., Ferreira, G.R., Arruda, H.F.D., Silva, F.N., Comin, C.H., Amancio, D.R., & Costa, L.F. (2021). Principal Component Analysis: A Natural Approach to Data Exploration. ACM computing surveys,  54 (4): 70:1–70:34.
Haute, S.V., Nikkhah, A., Malavi, D., & Kiani, S. (2023). Prediction of essential oil content in spearmint (Mentha spicata) via near-infrared hyperspectral imaging and chemometrics. Scientific Reports, 13(1), 4261.
Javadi, K., Ghasemi-Varnamkhasti, M., and Sabzi, N. (2017). Freshness detection of beef using image processing and response surface. Iranian Journal of Food Science and Technology Research, 13(2), 251-261.
Kiani, S., Rahimzadeh, H., Kalantari, D., & Moradi-Sadr, J. (2023). Aroma modeling and quality evaluation of spearmint (Mentha spicata subsp. spicata) using electronic nose technology coupled with artificial intelligence algorithms. Journal of Applied Research on Medicinal and Aromatic Plants, 35, 100473.
Koekoeh, S., Riana, P., Nugrahaeni, P. A., Ronald, T., & La Jumadin, P. (2022). Meat quality detection using imaging and AI analysis. In Proceedings of the 2nd International Conference on Research and Development (ICORAD), 41–48. https://doi.org/10.47841/icorad.v1i2.94
Ma, J., Sun, D. W., Pu, H., Wei, Q., & Wang, X. (2019). Protein content evaluation of processed pork meats based on a novel single shot (snapshot) hyperspectral imaging sensor. Journal of Food Engineering, 240, 207–213.
Meenu, M., Kurade, C., Neelapu, B.C., Kalra, S., Ramaswamy, H.S., & Yu, Y. (2021). A concise review on food quality assessment using digital image processing. Trends in Food Science & Technology, 118, 106-124.
Asadi, M.R., Kalantari Hessari, A., Ghaemmagami, S.S., Mosleh, N., Ghorbanzadeh, B., & Abdi., P. (2023). Meat products components using histological method and image analysis software. Veterinary Research and Biologic Products, 1, 102-112. DOI:  10.22092/vj.2022.358361.1966. Minaei, S., Kiani, S., Ayyari, M., & Ghasemi-Varnamkhasti, M. (2017). A portable computer-vision-based expert system for saffron color quality characterization. Journal of applied research on medicinal and aromatic plants, 7, 124-130.
Morsy, N, & Sun, D.W. (2013). Robust linear and non-linear models of NIR spectroscopy for detection and quantification of adulterants in fresh and frozen-thawed minced beef. Meat Science, 93 (2), 292-302.‏
Pernice, V., Staude, B., Cardanobile, S., & Rotter, S. (2011). How structure determines correlations in neuronal networks. Plos Computational Biology, 7(5), e1002059.‏
Rady, A. M., Adedeji, A., & Watson, N. J. (2021). Feasibility of utilizing color imaging and machine learning for adulteration detection in minced meat. Journal of Agriculture and Food Research, 6, 100251.
Sadeghpour, A., Pirzadeh Khalil, A., Sehatkhah., M., & Khandaghi, J. (2019). Detecting adulteration and checking the authenticity of meat and meat products using chemical properties and PCR technique in Tabriz city. Health and Hygiene of Ardabil, 11(4), 478-488.
Sampaio, P. S., Castanho, A., Almeida, A.S., Oliveira, J., & Brites, C. (2020). Identification of Rice Flour Types With Near‐Infrared Spectroscopy Associated With PLS‐DA and SVM Methods. European Food Research and Technology 246: 527–537.
Singh, P., Singh, N., Kant Singh, K., & Akansha, S. (2021). Chapter 5 - Diagnosing of disease using machine learning. Machine Learning and the Internet of Medical Things in Healthcare, 89-111. https://doi.org/10.1016/B978-0-12-821229-5.00003-3
Shen, C., Wang, R., Nawazish, H., Wang, B., Cai, K., Xu, B. (2024). Machine vision combined with deep learning–based approaches for food authentication: An integrative review and new insights. Comprehensive Review, 23(6), e70054. https://doi.org/10.1111/1541-4337.70054
Varshney, L.R., Chen, B.L., Paniagua, E., Hall, D.H., & Chklovskii, D.B. (2011). Structural properties of the Caenorhabditis elegans neuronal network. PLoS computational biology, 7(2), e1001066.‏
World Health Organization. (2024). Red meat consumption in Iran and the world. https://www.who.int
Yang, Q., Sun, D. W., & Cheng, W. (2017). Development of simplified models for nondestructive hyperspectral imaging monitoring of TVB-N contents in cured meat during drying process. Journal of Food Engineering, 192, 53–60.