Food Waste Detection

Learn more about related image processing and quantity estimation steps in the Quantity Estimation and Food Quantity Detection pages.

Plan & Token Requirements

Feature available in the following LogMeal Plans:

Accessible by the following User Types:

What It Does

Food Waste Estimation allows developers to calculate food leftovers by analyzing post‑meal images and comparing them to recognized pre-meal intakes.
It estimates the remaining portion per dish, computes waste percentage, and provides eaten vs. wasted amounts for nutritional tracking and sustainability reporting.

This feature can be optionally integrated with Quantity Estimation and the LogMeal Depth SDK, leveraging multi‑angle or depth‑based imagery for accurate leftover volume estimation.

✕

✕

When to Use It / Outcomes

• When you need to measure leftovers after a meal to determine actual food consumption.
• To quantify food waste for sustainability, restaurant analytics, or cost reporting.
• To compare pre‑meal and post‑meal quantities to compute eaten vs. wasted amounts.

Feature-Specific Details

• Works on registered intakes captured via food segmentation.
• Accepts either RGB sequences (multi-angle captures) or single RGB images devices non-compatible with LogMeal Depth SDK.
• Returns for each dish:
  • Leftover region mask and dish label with confidence.
  • Remaining quantity/volume estimated against pre‑meal serving size.
• Supports both single-dish and multi-dish plates.
• Integrates directly with Quantity Estimation to compute the initial quantity.
• Enables nutritional recalculation of eaten vs. wasted nutrients.
• Offers confirmation endpoint to validate and correct leftover data before storage.

Related Endpoints

Use the following endpoints for food waste estimation and confirmation:

• POST /v2/waste/detection/intake → 🔴 Detect leftovers (remaining food) from a registered intake using an RGB image or sequence.
• POST /v2/waste/confirm → 🔴 Confirm detected dishes and leftover quantities after waste detection.

Remember to review each endpoint’s specific request limitations.

Typical Workflow

1. Capture pre‑meal image using either Food Segmentation with Quantity Estimation or Regular Food Segmentation.
2. Run leftover detection on post-meal image using POST /v2/waste/detection/intake. Optionally use the LogMeal Depth SDK.
3. Review and confirm results via POST /v2/waste/confirm.
4. Analyze the final results in the user’s nutritional record, where you will find the leftovers, initial serving and real intake (subtracted).

Related Use Cases