From photo to estimate
From a photo, Volumeal identifies the foods present and estimates the most likely amount of each. It derives calories and macros — protein, carbs, fat. You can also scan a barcode or dictate your meal: three inputs, one logic.
The estimate is never shown as a single number. We display a range and a confidence level, because that's the honest shape of reality.
How we compute the interval
We combine visual recognition (foods, apparent volume) with nutritional composition databases. The uncertainty comes mostly from what a photo can't see: cooking oil, added sugar, real density. We translate it into an interval that covers the real value in the vast majority of cases, with a confidence level.
What tightens the range
Top-down or 3/4, whole plate in frame.
A bank card in frame gives real portion sizes.
Adjust the portion in one tap: the interval and confidence recompute.
Where the numbers come from
Nutritional values rely on recognised public food-composition databases, supplemented by label data for packaged products. When sources diverge, we keep a range rather than a single value — consistent with our range.
What a photo can't know
An image doesn't reveal everything: hidden fats, dissolved sugar, cooking method, ingredients under the surface. Volumeal estimates as best it can and expresses the rest as uncertainty — it never claims to-the-gram precision.
Volumeal is a tracking tool, not a medical device. Estimates don't replace the advice of a healthcare professional.