Food Waste Classification Using Image and Weight Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for classifying food items in commercial kitchens are inefficient, unsustainable, and prone to errors, particularly in environments without point-of-sale systems, leading to operational inefficiencies and poor waste management.
Innovation Solution
A system utilizing multiple sensors, including image and weight sensors, captures data during food item events and processes it using a model trained on historical data to automatically classify food items, enhancing accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual classification methods are used by food consultants or hierarchical menus, then classification can be performed, but the process is slow, labor-intensive, and not scalable
Solution Approach 1:
The patent replaces manual mechanical classification processes with an automated optical sensing system. Image sensors capture food waste images, and machine learning models automatically classify food items, eliminating the need for manual hierarchical menu navigation and significantly increasing classification speed while improving automation level.
Solution Approach 2:
The system enables self-service classification where the food waste itself provides the classification information through its visual characteristics. The image sensor and ML model work together to automatically identify and classify food items without requiring human intervention, making the system self-sufficient in the classification task.
2Loss of information
If multiple disposal events are accumulated in the waste receptacle, then continuous monitoring is possible, but isolating new food items from previous events becomes more difficult
Solution Approach 1:
The system performs preliminary actions by capturing images at each disposal event and storing them for later processing. When new food items are added, the system compares new images against previously captured images to identify changes, allowing accurate isolation of new food items without re-processing the entire waste receptacle contents.
Solution Approach 2:
The patent segments the image processing task by dividing the waste receptacle contents into discrete events. Each disposal event is captured as a separate image, and the system processes these segmented images individually, comparing only the new event against previous ones to identify new food items, thereby reducing processing complexity.
3Measurement precision
If automated image processing is implemented, then classification accuracy improves, but processing time and computational resources increase
Solution Approach 1:
The system applies partial action by processing only the necessary portion of image data. Instead of analyzing entire waste receptacle images in full detail, the ML model focuses on identifying and classifying only the new food items that have been added, reducing processing time while maintaining classification accuracy for the relevant portions.
Solution Approach 2:
The system uses periodic action by capturing images at discrete disposal events rather than continuously monitoring. This periodic sampling approach reduces the total processing time while maintaining accurate classification, as the system only needs to process images when new food items are introduced to the waste receptacle.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system enables automated and accurate classification of food items, reducing user errors and increasing operational efficiency in commercial kitchens.
Implementation Method 1
image data captured from an image sensor above the waste receptacle
Implementation Method 2
weight data captured from a weight sensor within a scale under the weight receptacle
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a method for classifying food items. The method includes the steps of: capturing one or more sensor data relating to a food item event; and classifying the food item, at least in part, automatically using a model trained on sensor data. A system and software are also disclosed.