We propose methods to automatically detect food item and estimate calorie from a given image of food item. Such model can then be ported to mobile devices, therefore it can serve as a way to automatically record calorie intake. Such system can be used in health-care industry to monitor the patients diet habits. Integrating such system in wearable devices such as Google Glass would further ease recognition and recoding of food items. Our goal was to provide a benchmark for computer vision researchers who are working on this area rather than to propose such system as state-of-the-art.
Fastfood7
Food Recognition and Calorie Extraction using Bag-of- SURF and Spatial Pyramid Matching Methods
Deep Learning-Based Food Calorie Estimation Method in Dietary Assessment