Digital Commons Network^™

Quantification Of Energy Intake Using Food Image Analysis, Sertan Kaya

LSU Master's Theses

Obtaining real-time and accurate estimates of energy intake while people reside in their natural environment is technically and methodologically challenging. The goal of this project is to estimate energy intake accurately in real-time and free-living conditions. In this study, we propose a computer vision based system to estimate energy intake based on food pictures taken and emailed by subjects participating in the experiment. The system introduces a reference card inclusion procedure, which is used for geometric and photometric corrections. Image classification and segmentation methods are also incorporated into the system to have fully-automated decision making.