Food Groups Detection

Learn more about how to preprocess your images in our Image Pre-processing Tutorial page.

Plan & Token Requirements

Feature available in the following LogMeal Plans:

Accessible by the following User Types:

What It Does

Food Groups Detection returns high-level food categories for detected dishes.
Example: a sushi prediction may return groups like fish and rice. This is useful for diet pattern analysis, pyramid-style reporting, and as a building block for features like Variety Score.

✕

✕

When to Use It / Outcomes

• You need broad dietary tags for a dish or an image (e.g., does it contain vegetables, egg, meat?).
• You want simple, interpretable labels for dashboards and reports (rather than full ingredient breakdown).
• Output: JSON containing one or more food group labels for each detected dish or for the image/dish type result.

Feature-Specific Details

Current groups (representative list):

• meat — white or red meat
• dessert — typical dessert dishes
• dairy products — milk, yogurt, etc.
• seafood — shellfish (clams, prawns, etc.)
• rice — rice or rice-containing dishes
• fruit — fresh or cooked fruit
• noodles/pasta — pasta & noodles
• vegetables — raw or cooked vegetables
• fish — fish & fish products
• bread — breads & grain-like products
• fried food — items cooked by frying
• egg — contains a relevant amount of egg
• soup — soup-like dishes
• cereal/grain — cereals or similar
• nuts — nuts & seeds
• alcohol — alcoholic beverages
• beverage — beverages (general)
• soda — sugared soft drinks
• tubers and derivatives — potatoes, yuca, etc.
• sauces and dressings — condiments
• legumes — beans, lentils, etc.
• whole grain — made with whole grains
• water — contributes to water intake

Notes

• Groups are tied to the recognized dish (i.e., see Several Dishes Recognition).
• Returned groups can be used downstream for diet labels, variety scoring, and recommendations.

Related Endpoints

In order to get the food groups for your image you must detect the dishes appearing on it:

• POST /image/segmentation/complete → 🔴 Detects all dishes in the image and returns predictions that include food groups per dish region.

Remember to check applicable request limitations inside each of the endpoints. These endpoints are available to 🔴 APIUser tokens and are included in Analyse plans or above.

Typical Workflow

1. Capture an image of the meal.
2. Call Segmentation Complete to get per-dish predictions (recommended when multiple items are present), or call Food Yype Recognition for a single view / pre-filter.
3. Read the food_groups (or equivalent field) from the prediction(s) and store them with the intake.
4. Optionally feed groups into reports (e.g., variety, food pyramid coverage) or recommendation pipelines.

Related Use Cases

Copy-Paste Recipes

Set profile data once to improve label language and regional dishes:

Send the image to segmentation to obtain regions and top dish candidates per region. Display the top 5 candidate dishes and ask the user for confirmation: