Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{199103,
        author = {Varad Digraskar and Manisha Pawar and Kedar Fulsawange and Prithviraj Gavhane and Shreya Jakhotiya},
        title = {NutriScan: AI-Powered Food Recognition and Nutrition Estimation with Applications in Fruit Grading},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {15262-15270},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=199103},
        abstract = {Being conscious of what we eat is necessary to maintain a healthy diet, but most people still struggle with accurate dietary tracking. Conventional techniques like manual food logging, calorie counting, or meal recall at the end of the day are labor-intensive, prone to mistakes, and frequently discontinued over time. Food photos offer a natural and practical source of information that can be used to streamline dietary assessment because smartphones and cameras are so widely available. This paper introduces NutriScan, a web-based, vision-based framework for estimating nutritional data from user-captured food photos. The suggested system enables a user to upload or take a picture of a meal using a browser. The system then uses image segmentation and food classification techniques to identify each individual food item on the plate. NutriScan takes a pragmatic approach by estimating portion sizes using common serving assumptions because accurate quantity measurement from a single image is not possible without depth information. The identified food items are then mapped to food composition databases to estimate nutritional values like calories and macronutrients. By using an optical character recognition (OCR) module to directly extract nutritional data from food labels, the framework supports packaged foods in addition to cooked or prepared meals. To produce understandable nutritional insights and warnings, the extracted data is normalized and analyzed using straightforward rule-based logic. The system is accessible without the need for a specific mobile application thanks to the overall design, which places an emphasis on low latency, transparency, and ease of use in a browser-based environment. This work's main contribution is the presentation of a methodological pipeline and a clear system architecture for an image-based nutrition estimation platform that is practical to use in the real world. This paper focuses on the design decisions, underlying methods, and assessment approach that can direct the NutriScan framework's future implementation and validation rather than presenting experimental results. NutriScan seeks to minimize the effort needed for dietary tracking while being transparent about estimation limitations by integrating food recognition, approximate portion reasoning, and explainable nutrition analysis into a single web framework.},
        keywords = {Food image recognition; Nutrition estimation; Dietary assessment; Computer vision; Instance segmentation; Optical character recognition (OCR); Web-based application; Approximate portion estimation},
        month = {April},
        }