NutriVision: AI-Powered Real-Time Food Recognition and Personalized Diet Planner

Abstract

NutriVision is an AI-powered mobile application designed to provide personalized diet planning and real-time food recognition. The application allows users to enter personal details such as height, weight, age, gender, allergies, medical conditions (such as diabetes), and dietary preferences. Based on this information, the system generates a weekly diet plan by delivering accurate nutritional insights for each food item. A key feature of NutriVision is its real-time food recognition capability, where users can capture images of their meals using a mobile camera. The system then identifies the food items instantly and provides detailed nutritional information, including calories, proteins, carbohydrates, and fats. It also generates alerts for potential allergens and warns users when detected food items may be unsuitable for specific health conditions such as diabetes.

Introduction

The text describes NutriVision, an AI-powered mobile application designed to improve dietary awareness and help users manage health conditions like obesity, diabetes, and malnutrition through smart nutrition tracking.

The system uses computer vision and machine learning to recognize food from images captured via a smartphone. It then provides detailed nutritional information such as calories, proteins, fats, and carbohydrates. Users also enter personal health data (age, weight, allergies, medical conditions), which is used to generate personalized diet plans and dietary recommendations using standard nutritional formulas like the Mifflin–St Jeor equation.

NutriVision includes key features such as:

• Real-time food recognition using CNN-based models
• Personalized weekly diet planning
• Nutrient gap analysis to detect deficiencies
• Allergy and diabetic alerts for unsafe foods
• Progress tracking through graphs and reports

The literature review highlights existing research in food recognition, deep learning models, and dietary tracking systems, showing that CNN-based approaches are effective but still face challenges like dataset limitations and complex food variations.

The system is designed with a layered architecture including user interface, AI processing, application logic, and database storage (e.g., Firebase/MongoDB). It processes images, identifies food items, estimates portions, and generates nutritional analysis in real time.

Conclusion

In conclusion, Nutrivison presents a modern and intelligent solution to one of the most pressing health challenges faced today—managing diet and nutrition in a personalized, accurate, and accessible way. By combining artificial intelligence, computer vision, and mobile technology, this system enables users to effortlessly track their food intake through image recognition, receive tailored diet plans based on their health conditions and preferences, and get real-time alerts for allergic or harmful food items. Unlike traditional diet tracking methods that rely on manual entry or generic recommendations, Nutrivison offers a highly personalized and automated approach that adapts to the user’s changing health data over time. The integration of deep learning for food identification, portion estimation, and nutritional analysis allows the system to deliver accurate results even in real-world scenarios involving mixed meals and varying portion sizes. Moreover, the system’s ability to analyze nutrient gaps, provide health-conscious meal suggestions, and generate weekly and monthly progress reports empowers users to make informed dietary choices and stay committed to their goals. The inclusion of features like diabetic alerts and allergy detection also adds a safety layer that makes the app more than just a diet tracker—it becomes a digital health assistant.

References

[1] C. Liu, Y. Li, L. Luo, S. E. George, and Y. Yao, \"DeepFood: Deep Learning-Based Food Image Recognition for Computer-Aided Dietary Assessment,\" arXiv preprint arXiv:1606.05675, 2016. [Online]. Available: https://arxiv.org/abs/1606.05675 [2] M. Hu, Y. Zheng, G. Li, Z. Li, and X. Yao, \"A Comprehensive Survey on Food Computing,\" IEEE Transactions on Knowledge and Data Engineering, vol. 33, no. 6, pp. 2322–2343, 2020. [Online]. Available: https://ieeexplore.ieee.org/document/9072256 [3] H. Aizawa and Y. Ogawa, \"FoodLog: Multimedia tool for capturing daily food intake,\" Proceedings of the 2nd ACM International Workshop on Multimedia for Personal Health and Health Care, pp. 23–26, 2015. [Online]. Available: https://ieeexplore.ieee.org/document/7047326 [4] A. Patil, N. Mehta, R. Kumar, and S. Mahajan, \"Smart Diet Planner: A Personalized Diet Recommendation System,\" International Journal for Research in Applied Science and Engineering Technology (IJRASET), vol. 8, no. 7, pp. 1012–1017, 2020. [Online]. Available: https://www.ijraset.com/research-paper/smart-diet-planner [5] G. Jadhav, S. Kawade, A. Sonawane, and P. Gole, \"NutriScan: AI-Based Ingredient Detection and Evaluation,\" International Journal of Engineering Research & Technology (IJERT), vol. 14, no. 5, pp. 320–324, May 2023. [Online]. Available: https://www.ijert.org/research/nutriscan-ai-based-ingredient-detection-and-evaluation-IJERTV14IS050197.pdf [6] M. Bossard, L. Guillaumes, C. Llobet, and X. Giro-i-Nieto, \"Food-101 – Mining Discriminative Components with Random Forests,\" European Conference on Computer Vision (ECCV), 2014. Dataset Survey Source: ACM Computing Surveys, 2018. [Online]. Available: https://dl.acm.org/doi/10.1145/3193122

Copyright

Copyright © 2026 Ms. H. R. Agashe , Harshada Karad, Divya Karanjkar , Sayali Bhosale, Renuka Chitte. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.