Vitamin deficiencies pose significant health risks if left undiagnosed, particularly in regions with limited access to medical testing. This research presents a dl based system for automated detection of vitamin deficiencies using image classification. Leveraging the MobileNetV2 (Yashvi Chandola, 2021) architecture, the model is trained to classify images into six categories corresponding to deficiencies in Vitamins A, B, C, D, E, and K.
The system utilizes TensorFlow and a transfer learning approach to enhance training efficiency and accuracy. A user-friendly graphical interface built with Tkinter allows users to upload images and receive real-time predictions, accompanied by detailed descriptions of symptoms, dietary sources, and medical advice for the detected vitamin deficiency. Experimental results indicate promising accuracy of 98% across multiple classes, supporting the feasibility of deploying this lightweight application in clinical field settings. The proposed system also introduces additional features such as confidence-based prediction display, rich nutritional advice, and expandable class support, making it a practical tool for early nutritional screening and public health education.
Jetzt kaufen
- Arbeit zitieren
- Deva Kranthi (Autor:in), 2026, Vitamin Deficiency Detection using Deep Learning, München, GRIN Verlag, https://www.grin.com/document/1737355
