A video-based algorithm for food intake estimation in the study of obesity

A lack of appropriate tools to evaluate food consumption is currently a significant obstacle in identifying the most important factors that contribute to obesity. We show that by using a wearable video camera to chronically record an individual’s daily activities, food intake can be assessed more accurately. In this paper we present a novel and simple video-based algorithm to objectively estimate the physical dimensions of objects on top of a table. The intention is to use such an algorithm in the future to estimate the dimensions and quantities of various food products. The proposed algorithm has shown superior accuracy to subjective estimations.