Vitamin deficiencies can manifest through visually distinct symptoms and indications across various parts of the human body. The application offers individuals the ability to identify potential vitamin deficiencies without requiring blood samples. By analyzing photos of their eyes, lips, tongue, and nails, users can detect various visually distinct symptoms and signs that may indicate a wide range of vitamin deficiencies affecting different parts of the body. Here, we have considered the images dataset of eyes, nails, tongue and lips. Initially, the preprocessing is performed to the data and then CNN and ANN algorithms are used to train the data and tested by using OpenCV. Transfer learning is also made use, using pre-trained models such as MobileNet and NASNet to enhance the performance of the model. The neural network is trained on an image dataset that contains images of Vitamin deficiency and healthy images. To validate the proposed approach, we conduct experiments on a large dataset, assessing the model’s accuracy, sensitivity, and performance. We compare our results with existing methods and demonstrate the effectiveness of our approach in accurately detecting vitamin deficiencies. This model emphasizes an existing method that which was designed using the some of the algorithms of machine learning, which was not performed accurately and classification is not up to the expected level accuracy.