SMART AI INGREDIENT ANALYZER

This paper presented the “Smart AI Ingredient Analyzer,” an intelligent and accessible AI-powered system for real-time food ingredient analysis and health-risk prediction. By integrating OCR techniques, AI/ML models such as CNN, Decision Trees, Random Forest, and Logistic Regression, along with food databases, medical records, and knowledge graphs, the system provides accurate detection of allergens, harmful additives, and drug–food interactions. The personalized recommendation feature enables safer food suggestions based on user medical profiles, improving preventive healthcare and smart nutrition management. The user-friendly interface, multilingual support, and voice-enabled assistance further enhance accessibility and practical usability  without  requiring  specialized  hardware. Future work will focus on improving OCR accuracy for low-quality and multilingual food labels, expanding food and medical datasets, and integrating advanced deep learning models for more precise risk prediction. We also plan to enhance real-time cloud synchronization, IoT-based food quality monitoring, and intelligent voice interaction to create a more scalable, adaptive, and healthcare-oriented food safety ecosystem.

[1]          R. Watanabe, “Ingredient Detection from Low-Quality Food Label Images Using Deep Learning,” IEEE Access, vol. 13, pp. 10234–10245, 2025. [2]          M. M. Rahman, S. Ahmed, and T. Hasan, “Graph Mining for Food–Drug Interaction Prediction,” Elsevier Journal of Biomedical Informatics, vol. 128, 2022. [3]          H. Luo, Y. Zhang, and X. Li, “Deep Learning and Knowledge Graph-Based Drug–Drug Interaction Prediction,” IEEE Transactions on Artificial Intelligence, vol. 4, no. 2, pp. 245–256, 2023. [4]          M. Nagayi et al., “OCR Performance Analysis on Food Packaging Labels,” International Journal of Computer Vision and Applications, vol. 15, no. 3, pp. 120–130, 2021. [5]          M. Wang et al., “Drug–Drug Interaction Prediction Using Knowledge Graph and Graph Embedding,” BMC Bioinformatics, vol. 22, no. 1, pp. 1–15, 2021. [6]          X. Lin, J. Chen, and H. Zhao, “Deep Learning and Graph Learning Approaches for Drug Interaction Prediction,” Springer Nature Computer Science, vol. 4, no. 7, pp. 1–14, 2023. [7]          S. Ibrahim, A. Khan, and P. Roy, “OCR and Machine Learning for Nutritional Claim Verification on Food Labels,” IEEE Access, vol. 12, pp. 45871–45884, 2024. [8]          USDA FoodData Central, “FoodData Central Database.” https://fdc.nal.usda.gov/ [9]          Open Food Facts Contributors, “Open Food Facts Database.” https://world.openfoodfacts.org/ [10]     DrugBank Online, “Drug and Food Interaction Database.” https://go.drugbank.com/ [11]    TensorFlow Developers,“TensorFlow Documentation.” https://www.tensorflow.org/ [12]     Flask Documentation, “Flask Web Framework.” https://flask.palletsprojects.com/