The idea is to help school officials create healthier lunches based on what kids actually like to eat. The data can also help parents plan better meals at home: if Junior’s eating nothing but French fries at lunchtime, parents can try to balance out his diet with a dinner of vegetable and lean proteins.

Epidemiologists who focus on childhood nutrition at the public-health level could also use the findings to determine where the most pressing changes need to be made — and to better understand how children’s diets factor into rates of obesity and weight-related diseases like Type 2 diabetes…

The cameras take snapshots of children’s trays before they sit down and again after they’ve eaten; using sophisticated food-recognition software, the computer program analyzes the color, shape, volume, density and texture of every single piece of food on a child’s tray and compares them to a database of 7,500 foods. That allows researchers to monitor the calories, fat, fiber, sugar and protein — a total of 128 nutrients — kids eat in each meal.

Although each child’s meals will be tracked by a bar code on his or her tray, the cameras will not capture images of the kids themselves.