Food preparation method and system based on ingredient recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional food preparation systems rely on manual inputs for cooking temperatures and durations, are inaccurate in ingredient recognition, computationally intensive, and not easily extensible, limiting their ability to provide consistent and scalable food preparation controls and nutritional recommendations.
Innovation Solution
A food preparation system with in situ image acquisition and processing, using a two-stage classification method (general and detailed) to recognize ingredients, reducing computational complexity and improving accuracy, and adjusting heating units based on recognized ingredients for optimal cooking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If automatic ingredient recognition is implemented using AI and deep learning techniques, then ingredient recognition capability is improved, but computational resources required and system complexity increase significantly
Solution Approach 1:
The patent segments the ingredient recognition task into two distinct stages: a first classification model that identifies coarse categories (e.g., meat, vegetables, grains) and a second classification model that identifies specific ingredients within each category. This segmentation reduces the computational burden of each individual model and improves overall system efficiency while maintaining high recognition accuracy.
2Adaptability or versatility
If a comprehensive database of hundreds or thousands of food ingredients is included in the recognition model, then recognition coverage is improved, but model size and computational requirements increase
Solution Approach 1:
The patent divides the comprehensive ingredient database into multiple hierarchical levels. The first classification model handles coarse-grained categories covering hundreds of ingredient types, while the second classification model handles fine-grained specific ingredients within each category. This hierarchical segmentation allows the system to maintain comprehensive coverage without requiring a single monolithic model, thereby reducing individual model sizes and computational requirements.
3Adaptability or versatility
If image processing is performed under varied capture conditions, then system adaptability to different environments is improved, but recognition accuracy decreases due to false results
Solution Approach 1:
The patent performs preliminary image preprocessing operations including background removal, normalization, and enhancement before feeding images to the classification models. By standardizing the input images in advance, the system maintains high recognition accuracy across varied capture conditions without requiring the models to learn from noisy, inconsistent data.
4Measurement precision
If manual input of ingredient names and quantities is required for nutritional information, then nutritional accuracy is improved, but user convenience and system efficiency deteriorate
Solution Approach 1:
The patent implements an automated ingredient recognition system that captures images of food items and automatically identifies ingredients and estimates quantities using computer vision and classification models. This self-service approach eliminates the need for manual ingredient entry while maintaining acceptable accuracy for nutritional calculations, significantly improving user convenience and system efficiency.
Data Source
Figure 1
Figure 2
Figure 3(a)~3(c)
AI summary
A food preparation system and method include: triggering image capturing of a camera to obtain one or more images of a food support platform while the food support platform supports a first food item; performing ingredient recognition for the first food item, including: classifying a feature tensor of a respective image in a general classifier to identify one or more first-level food ingredient categories corresponding to the first food item; and classifying the feature tensor of the respective image in a respective detailed classifier corresponding to each of the one or more first-level food ingredient categories to identify a corresponding second-level food ingredient category corresponding to the first food item, wherein the second-level food ingredient category is a sub-category of said each first-level food ingredient category; and, adjusting one or more heating units for heating the first food item in accordance with the ingredient recognition that has been performed.