Smartphone app could help health and diet conscious track caloric intake by taking a picture
February 14, 2012
Purdue adjunct professor Carol Boushey reviews a software application that can identify the nutritional value in food by taking a picture of it with a smartphone. The application provides caloric numbers and the amount of protein, fat and carbohydrates in food through the Technology Assisted Dietary Assessment system, or TADA, being developed by Purdue's colleges of Health and Human Sciences and School of Electrical and Computer Engineering. (Purdue Research Foundation photo)
WEST LAFAYETTE, Ind. – A picture can be worth a thousand calories or 800 or 1,800 through a new smartphone application being developed at Purdue University to help the health and diet conscious track their caloric intake by taking a picture of their food.
The application counts more than calories. It also provides information on the amount of protein, fat and carbohydrates in food through the Technology Assisted Dietary Assessment system, or TADA, being developed by Purdue's College of Health and Human Sciences and School of Electrical and Computer Engineering.
"Our goal is to allow people to electronically record their food intake and help individuals with health challenges like diabetes understand what they're eating and make healthier choices," said Carol Boushey, who led the development team while a professor in Purdue's College of Health and Human Sciences. "Many people don't realize how much food and how many calories they are consuming, and that is one of the biggest contributors to the increasing weight we have in the United States and around the globe."
Boushey, who is now an adjunct professor at Purdue and director of the University of Hawaii Cancer Center's Nutrition Support Shared Resource, said the application's development entailed an interdisciplinary approach.
A smartphone user snaps a picture of a plate of food and drink. When the image is sent to Purdue's TADA software system, the user will receive the nutritional value and caloric count of the pictured items. (Purdue Research Foundation photo)
Edward J. Delp, the Charles William Harrison Distinguished Professor of Electrical and Computer Engineering, oversaw the imaging software development that automatically identifies food in the images. Marc Bosch-Ruiz and Margaret Zhu. graduate students in Delp's lab, created the software to determine a food's identity. Graduate student Junghoon Chae, and Ross Maciejewski, a visiting assistant professor, also worked on the application in professor David Ebert's lab to estimate the amount or volume of the food.
"We have excellent methods to measure physical activity, but we don't have a good way to measure what we consume," Delp said. "It's one thing for someone to type in a list of the food they've consumed and determine its nutritional value, but another to just take an image of a plate of food and anticipate that the application they are using can identify the type of food, the amount of food and then provide nutritional value."
Martin Okos, professor in the Department of Agricultural and Biological Engineering, and Shivangi Kelkar and Scott Stella, graduate students in Okos' lab, oversaw the food density application.
The Technology Assisted Dietary Assessment system, or TADA, could provide the nutritional value and caloric count of food and drink just by taking a picture with a smartphone. The application is being developed at Purdue University and commercialized through Purdue's Office of Technology Commercialization. (Purdue Research Foundation photo)
"The density of food is needed to convert the volume of food consumed to gram weight so that the app can determine nutritional value," Okos said. "To fill the app database, we are developing novel methods to accurately measure the density of all types of foods directly from a X-ray CT scan to produce a 3-D image and then confirmed the results with a laser scanner."
Bruce A. Craig, professor and director of the statistical consulting service for the Purdue College of Science's Department of Statistics, is also involved with the project. Funding for the project came from the National Institutes of Health.
The patented application has been tested with adolescents and adults and is being commercialized through Purdue's Office of Technology Commercialization. Contact Michael Halbrook, senior project manager, at 765-588-3483, firstname.lastname@example.org, for information.