AI-Based Meal Consumption Tracker
Budget: $5,000 – $10,000 SGD
1. Objective
To develop a system that allows patients to track food consumption by capturing before-and-after images of their meals. The system will use AI to estimate the percentage of food consumed with a margin of error not exceeding 5%.
2. Functional Requirements
2.1 Image Capture
The system shall allow the patient to take a photo before eating using a mobile or tablet device.
The system shall prompt the patient to take a second photo after eating from a similar angle and distance.
Image capture must follow standardized guidelines (e.g., top-down view, consistent lighting, fixed plate or tray) to ensure accuracy.
2.2 Image Comparison & Processing
The system shall process the before and after images using computer vision techniques.
AI shall be used to identify and segment food items within both images.
The system shall align and normalize the images for accurate comparison (adjust for lighting, position, and scale).
2.3 Consumption Estimation
The AI shall calculate the percentage of food consumed per item based on visible differences between the two images.
The system shall provide an overall consumption percentage.
The system shall maintain a margin of error not exceeding ±5% in consumption estimation.
2.4 Output & Reporting
The system shall display the consumption percentage in a clear visual format.
A summary log entry shall be generated after each meal (e.g., “85% of meal consumed; rice 90%, vegetables 70%, protein 100%”).
The system shall store image pairs and consumption records securely for dietitian review.
3. Non-Functional Requirements
Accuracy: System must maintain ≤5% error rate in estimating food volume consumed.
Performance: Image analysis and consumption report must be completed within 10 seconds of image submission.
Security: All patient data, including images and reports, must be encrypted and stored securely.
Usability: The interface must be intuitive and accessible to non-tech-savvy users (e.g., elderly patients).
Scalability: The system should support integration into hospital dietary tracking or wellness monitoring platforms.
To develop a system that allows patients to track food consumption by capturing before-and-after images of their meals. The system will use AI to estimate the percentage of food consumed with a margin of error not exceeding 5%.
2. Functional Requirements
2.1 Image Capture
The system shall allow the patient to take a photo before eating using a mobile or tablet device.
The system shall prompt the patient to take a second photo after eating from a similar angle and distance.
Image capture must follow standardized guidelines (e.g., top-down view, consistent lighting, fixed plate or tray) to ensure accuracy.
2.2 Image Comparison & Processing
The system shall process the before and after images using computer vision techniques.
AI shall be used to identify and segment food items within both images.
The system shall align and normalize the images for accurate comparison (adjust for lighting, position, and scale).
2.3 Consumption Estimation
The AI shall calculate the percentage of food consumed per item based on visible differences between the two images.
The system shall provide an overall consumption percentage.
The system shall maintain a margin of error not exceeding ±5% in consumption estimation.
2.4 Output & Reporting
The system shall display the consumption percentage in a clear visual format.
A summary log entry shall be generated after each meal (e.g., “85% of meal consumed; rice 90%, vegetables 70%, protein 100%”).
The system shall store image pairs and consumption records securely for dietitian review.
3. Non-Functional Requirements
Accuracy: System must maintain ≤5% error rate in estimating food volume consumed.
Performance: Image analysis and consumption report must be completed within 10 seconds of image submission.
Security: All patient data, including images and reports, must be encrypted and stored securely.
Usability: The interface must be intuitive and accessible to non-tech-savvy users (e.g., elderly patients).
Scalability: The system should support integration into hospital dietary tracking or wellness monitoring platforms.