Prediction of Healthy and Unhealthy Food Items using Deep Learning

In recent times, people are mostly attracted to fast food. In this paper, a model for categorizing food products using convolutional neural networks has been built. Deep learning facts are being used in day-to-day life and recognizing objects through image processing. Convolutional neural networks are effective for classifying photographs of food because they exclude distracting characteristics from images. By implementing this model, the layers involved here to convert the three-dimensional input into an output volume and to filter input to a higher level of abstraction. This proposed model may help people to identify healthy and unhealthy food items and save us from causes of diseases. This study used image datasets we built, with the images coming from social media and already-existing image datasets (Kaggle) and (Google). The findings of this study demonstrate how much more effective the system is in classifying images when compared to other approaches. The suggested method includes more visual elements than other method and that it performs better at classifying food images. According to the experimental findings, the classification of foods as healthy or unhealthy is 90% accurate.