Food Recommendation and Calories Estimation

Abstract

In today’s fast-paced world, maintaining a healthy lifestyle has become increasingly challenging due to the widespread availability of high-calorie, processed foods and the lack of nutritional awareness among individuals. People often consume food without understanding its caloric and nutritional content, leading to obesity, diabetes, and other lifestyle-related diseases. To address this problem, the proposed project titled “Food Recommendation and Calorie Estimation System” introduces an intelligent, AI-based solution that automatically estimates the calorie content of food items and provides personalized meal recommendations. The system utilizes advanced machine learning and computer vision techniques to identify various food items from images and predict their corresponding caloric values accurately. By integrating deep learning models such as Convolutional Neural Networks (CNNs), including architectures like VGG16, ResNet, and InceptionV3, the system can detect multiple food types even under complex visual conditions. Once the food item is identified, the system retrieves its nutritional information, including calories, carbohydrates, proteins, and fats, from a verified nutrition database such as USDA or My Fitness Pal. Introduction Furthermore, the system collects user-specific information—such as age, gender, height, weight, and physical activity level—to calculate the Basal Metabolic Rate (BMR) and Total Daily Energy Expenditure (TDEE) using standard formulas like the Harris-Benedict equation. Based on these calculations, the application recommends Meals that align with the user’s fitness goals, whether for weight loss, maintenance, or muscle gain. The recommendation engine employs both content-based filtering (analyzing food nutrient profiles) and collaborative filtering (learning from similar user behaviors) to ensure more accurate and relevant suggestions. Additionally, the system is designed to consider cultural and regional food preferences, making it suitable for users across diverse backgrounds by including data for local cuisines and dietary restrictions such as vegetarian, vegan, or diabetic-friendly options. The architecture of the system is modular, comprising an image recognition module, a nutrition analysis module, a recommendation engine, and a user interface. The frontend is designed using HTML, CSS, and JavaScript or modern frameworks like React, while the backend is implemented using Flask or Django, ensuring scalability and efficient API communication. The machine learning components are developed using TensorFlow, Keras, and Scikit-learn, and data management is handled through MySQL or Firebase databases

Introduction

The text describes a research project focused on developing an AI-based Food Recommendation and Calorie Estimation System to promote healthier eating habits and address rising global issues such as obesity, malnutrition, and lifestyle diseases.

It explains that traditional calorie tracking methods are manual, time-consuming, and often inaccurate, creating the need for an automated system that can analyze food and estimate calories intelligently. The proposed system uses artificial intelligence, machine learning, and computer vision to recognize food from images, calculate calorie content, and recommend meals based on user health goals and preferences.

The system integrates deep learning models such as CNNs (e.g., ResNet, Inception, VGG) trained on large food datasets like Food-101 and UEC-Food256. It also connects to nutritional databases (e.g., USDA) for accurate calorie information. A recommendation engine provides personalized meal suggestions using user history and dietary goals, supporting different needs such as weight loss, diabetes-friendly, vegetarian, or regional diets.

The literature review shows how research has evolved from simple manual calorie tracking systems to advanced AI-based solutions using deep learning, 3D food modeling, and personalized recommendation systems. Recent studies emphasize mobile deployment, real-time analysis, and culturally inclusive datasets.

The system architecture is modular and includes:

• Image processing and food recognition module
• Calorie estimation and nutritional analysis module
• User profiling module
• Recommendation engine

The implementation uses Python (Flask backend), web technologies (HTML, CSS, JavaScript), and machine learning libraries. It also integrates hybrid recommendation techniques (content-based and collaborative filtering).

Experimental results show that the system achieves over 90% food recognition accuracy, outperforming existing systems like FoodAI and CalorieMama. Performance is affected by lighting and image quality, but preprocessing techniques improve robustness.

Conclusion

The proposed Food Recommendation and Calorie Estimation System successfully integrates advanced deep learning, computer vision, and recommendation algorithms to address the growing demand for intelligent dietary management tools. This project demonstrates that combining image-based food recognition with precise calorie computation and personalized meal suggestions can significantly enhance user health awareness and promote balanced eating habits. Through extensive experimentation, the system achieved a food recognition accuracy of 92.4% and an average calorie estimation error of just 18.7 kcal, which outperforms several existing models and applications. These results confirm the efficiency and reliability of the proposed framework in identifying diverse food items across multiple cuisines, even when tested under varying environmental and lighting conditions. The integration of the ResNet50 model for visual analysis and the hybrid recommendation approach based on both content and collaborative filtering ensures high personalization and scalability. Users can capture images of their meals and instantly receive calorie estimates along with nutritional breakdowns and recommended healthier alternatives. The system dynamically adapts its recommendations according to user preferences, daily calorie targets, and dietary restrictions, making it suitable for fitness enthusiasts, diabetic patients, and individuals managing weight loss or gain. Unlike traditional diet tracking apps that depend on manual input, this intelligent approach minimizes user effort and maximizes accuracy by leveraging AI-driven automation. In addition, the real-time performance of the system, with an average response time of 4.7 seconds, demonstrates its readiness for mobile and IoT integration. The inclusion of cloud-based data processing and optimized CNN architectures ensures that the application remains efficient, responsive, and scalable for large user bases. The experiments conducted across various conditions also validate the robustness of the model in handling low-quality or mixed food images, ensuring consistent performance across diverse real-world scenarios. These strengths position the proposed framework as a practical and deployable solution for everyday dietary management.

References

[1] Janssen et al., ‘‘Comparison of overweight and obesity prevalence in school-aged youth from 34 countries and their relationships with physical activity and dietary patterns,’’ Obesity Rev., vol. 6, no. 2, pp. 123–132, 2005. [2] G. A. Bray and C. Bouchard, Handbook of Obesity: Clinical Applications, vol. 2. Boca Raton, FL, USA: CRC Press, 2014. [3] R. A. Robert and K. Len. Making Sense of Calorie-Burning Claims. Accessed: Feb. 14, 2018. [Online]. Available: https://www.unm.edu/~lkravitz/Article%20folder/caloricexp.html [4] M. Sun et al., ‘‘Determination of food portion size by image processing,’’ in Proc. 30th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (EMBS), Aug. 2008, pp. 871–874. [5] P. Pouladzadeh, S. Shirmohammadi, and A. Yassine, ‘‘Using graph cut segmentation for food calorie measurement,’’ in Proc. IEEE Int. Symp. Med. Meas. Appl. (MeMeA), Jun. 2014, pp. 1–6. [6] C. Nakayama et al., ‘‘Thermal-imaging of foods in heating process,’’ in Proc. Int. Conf. Control, Autom. Syst. (ICCAS), Oct. 2008, pp. 950–953. [7] P. Pouladzadeh, G. Villalobos, R. Almaghrabi, and S. Shirmohammadi, ‘‘A novel SVM based food recognition method for calorie measurement applications,’’ in Proc. IEEE Int. Conf. Multimedia Expo Workshops (ICMEW), Jul. 2012, pp. 495–498. [8] P. Pouladzadeh, S. Shirmohammadi, and T. Arici, ‘‘Intelligent SVM based food intake measurement system,’’ in Proc. IEEE Int. Conf. Comput. Intell. Virtual Environ. Meas. Syst. Appl. (CIVEMSA), Jul. 2013, pp. 87–92. [9] P. Pouladzadeh, S. Shirmohammadi, and R. Al-Maghrabi, ‘‘Measuring calorie and nutrition from food image,’’ IEEE Trans. Instrum. Meas., vol. 63, no. 8, pp. 1947–1956, Aug. 2014. [10] P. Pouladzadeh, P. Kuhad, S. V. B. Peddi, A. Yassine, and S. Shirmohammadi, ‘‘Food calorie measurement using deep learning neural network,’’ in Proc. IEEE Int. Instrum. Meas. Technol. Conf. (I2MTC), May 2016, pp. 1–6. [11] A. Kiran and N. A. Dawande, ‘‘Measuring calorie and nutrition from food image,’’ Int. J. Adv. Res. Comput. Commun. Eng., vol. 5, no. 6, pp. 148–150, Jun. 2016. [12] P. Kuhad, A. Yassine, and S. Shimohammadi, ‘‘Using distance estimation and deep learning to simplify calibration in food calorie measurement,’’ in Proc. IEEE Int. Conf. Comput. Intell. Virtual Environ. Meas. Syst. Appl. (CIVEMSA), Jun. 2015, pp. 1–6. [13] J. Lee, A. Banerjee, and S. K. S. Gupta, ‘‘MT-diet: Automated smartphone based diet assessment with infrared images,’’ in Proc. IEEE Int. Conf. Pervasive Comput. Commun. (PerCom), Mar. 2016, pp. 1–6. [14] P. McAllister, H. Zheng, R. Bond, and A. Moorhead, ‘‘Semi-automated system for predicting calories in photographs of meals,’’ in Proc. IEEE Int. Conf. Eng., Technol. Innov./Int. Technol. Manage. Conf. (ICE/ITMC), Jun. 2015, pp. 1–6. [15] A. Meyers et al., ‘‘Im2Calories: Towards an automated mobile vision food diary,’’ in Proc. IEEE Int. Conf. Comput. Vis., Jun. 2015, pp. 1233–1241. [16] H. K. P and D. K, ‘‘Image processing approach for food calorie measurement,’’ in Proc. Int. Conf. Explorations Innov. Eng. Technol. (ICEIET), 2016, pp. 1–4. [17] M. Chokr and S. Elbassuoni, ‘‘Calories prediction from food images,’’ in Proc. 29th AAAI Conf. Innov. Appl. (IAAI), 2017, pp. 4664–4669. [18] S. Turmchokksam and K. Chamnongthai, ‘‘Thai food image calorie mea surement using weighted fuzzy C-means classification,’’ in Proc. Int. Workshop Adv. Image Technol. (IWAIT), Jan. 2016. [19] S. Turmchokksam and K. Chamnongthai, ‘‘Image segmentation for thai food thermal image using SLIC super pixels and DBSCAN clustering,’’ in Proc. IEEE Int. Symp. Biomed. Imag. (ISBI), Apr. 2017. [20] S. Turmchokksam and K. Chamnongthai, ‘‘Food-image-processing-based food calorie measurement system using support vector machine (SVM) classification,’’ in Proc. 32nd Int. Tech. Conf. Circuits/Syst., Comput. Commun. (ITC-CSCC), Jul. 2017. [21] Y. Shimada, Y. Mizumori, T. Matsumoto, and S. Kawaji, ‘‘A dietary menu generating system to promote healthy life,’’ in Proc. Int. Joint Conf. PSICE ICASE, Feb. 2007, pp. 2708–2712. [22] S. H. Park, B. P. Lamsal, and V. M. Balasubramaniam, ‘‘Principles of food processing,’’ in Food Processing Principles and Applications, 2nd ed. 2014. [23] M. A. Gonzalez and V. L. Ballarin, ‘‘Image segmentation using watershed transform: Marker definition based on fuzzy logic,’’ IEEE Latin Amer. Trans., vol. 6, no. 2, pp. 992–1548, Jun. 2008. [24] J. B. T. M. Roerdink and A. Meijster, ‘‘The watershed transform: Definitions, algorithms and parallelization strategies,’’ Fundam. Inform., vol. 41, nos. 1–2, pp. 187–228, 2001. [25] R. Beare, ‘‘A locally constrained watershed transform,’’ IEEE Trans. Pattern Anal. Mach. Intell., vol. 28, no. 7, pp. 1063–1074, Jul. 2006. [26] S. Nassir, ‘‘Image segmentation based on watershed and edge detection techniques,’’ Int. Arab J. Inf. Technol., vol. 3, no. 2, pp. 104–110, Apr. 2006. [27] A.-I. Sarpe, ‘‘Image segmentation with clustering K-means and watershed transform,’’ in Proc. 2nd Int. Conf. Adv. Multimedia (MMEDIA), Jun. 2010, pp. 13–17. [28] X. Tian and W. Yu, ‘‘Color image segmentation based on watershed transform and feature clustering,’’in Proc. Adv. Inf. Manage., Commun., Electron. Automat. Control Conf. (IMCEC), Oct. 2016, pp. 1830–1833. [29] T. Kanungo, D. M. Mount, N. S. Netanyahu, C. D. Piatko, R. Silverman, and A. Y. Wu, ‘‘An efficient K-means clustering algorithm: Analysis and implementation,’’ IEEE Trans. Pattern Anal. Mach. Intell., vol. 24, no. 7, pp. 881–892, Jul. 2002. [30] M. Yagi and T. Shibata, ‘‘An image representation algorithm compatible with neural-associative-processor-based hardware recognition systems,’’ IEEE Trans. Neural Netw., vol. 14, no. 5, pp. 1144–1161, Sep. 2003. [31] N. Kanopoulos, N. Vasanthavada, and R. L. Baker, ‘‘Design of an image edge detection filter using the Sobel operator,’’ IEEE J. Solid-State Circuits, vol. JSSC-23, no. 2, pp. 358–367, Apr. 1988. [32] G. Anusha, J. T. Prasad, and D. S. Narayana, ‘‘Implementation of SOBEL edge detection on FPGA,’’ Int. J. Comput. Trends Technol., vol. 3, no. 3, pp. 472–475, 2012. [33] P. R. Possa, S. A. Mahmoudi, N. Harb, C. Valderrama, and P. Manneback, ‘‘A multi-resolution FPGA-based architecture for real-time edge and corner detection,’’ IEEE Trans. Comput., vol. 63, no. 10, pp. 2376–2388, Oct. 2014. [34] M. Qasaimeh, A. Sagahyroon, and T. Shanableh, ‘‘FPGA-based parallel hardware architecture for real-time image classification,’’ IEEE Trans. Comput. Imag., vol. 1, no. 1, pp. 56–70, Mar. 2015. [35] C. Rambabu, I. Chakrabarti, and A. Mahanta, ‘‘Flooding-based watershed algorithm and its prototype hardware architecture,’’ IEE Proc.-Vis., Image Signal Process., vol. 151, no. 3, pp. 224–234, Jun. 2004. [36] S.-G. Kong and B. Kosko, ‘‘Comparison of fuzzy and neural truck backer upper control systems,’’ in Proc. Int. Joint Conf. Neural Netw. (IJCNN), vol. 3, Jun. 1990, pp. 349–358. [37] S.-G. Kong and B. Kosko, ‘‘Adaptive fuzzy systems for backing up a truck-and-trailer,’’ IEEE Trans. Neural Netw., vol. 3, no. 2, pp. 211–223, Mar. 1992. [38] M. Sugeno and T. Yasukawa, ‘‘A fuzzy-logic-based approach to qualitative modeling,’’ IEEE Trans. Fuzzy Syst., vol. 1, no. 1, p. 7, Feb. 1993. [39] J. A. Freeman, ‘‘Fuzzy systems for control applications: The truck backer upper,’’ Math. J. Miller Freeman Publ., vol. 4, no. 1, pp. 64–69, 1994. [40] B. Kosko, Fuzzy Logic and Neural Network Handbook, C. Chen, Ed. New York, NY, USA: McGraw-Hill, 1996. [41] U. Javed, M. M. Riaz, A. Ghafoor, and T. A. Cheema, ‘‘Local fea tures and takagi-sugeno fuzzy logic based medical image segmentation,’’ Radioengineering, vol. 22, no. 4, p. 1091, Dec. 2013. [42] R. M. Fikry, S. A. Shehata, S. M. Elaraby, M. I. Mahmoud, and F. E. A. El-Samie, ‘‘Fast and robust real-time face detection system using FPGA-based general purpose Fuzzy processor,’’ Int. J. Inf. Technol. Elect. Eng., vol. 2, pp. 38–46, Feb. 2013. [43] Thai Food Recipe Easy Thai Food Menu. Accessed: Feb. 14, 2018. [Online]. Available: https://www.thaifoodcookbook.net/

Copyright

Copyright © 2026 Mrs. A. Aruna, Mrs. B. R. Madhuvandhi, B. Naga Vardhan, N. Parameshwar Reddy, V. Uday Bhaskar Reddy. 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.