Develop an AI-assisted image analysis model for food waste composting
Project Description
We are converting campus food waste into useful organic fertilizer using a rapid food waste composting system. The project aims to explore the use of machine learning and computer vision techniques to analyze images of food waste and composted fertilizer, and correlate visual features with experimental measurements and composting performance.
Supervisor
LIU Yuanshuai
Quota
2
Course type
UROP1000
UROP1100
UROP2100
UROP3100
UROP3200
UROP4100
Applicant's Roles
Conduct image collection of food waste and composted fertilizer samples under controlled conditions.
Assist in organizing, labeling, and managing image datasets for machine learning model development.
Assist in developing and testing machine learning or computer vision models for image-based analysis of food waste composition and fertilizer quality.
Analyze experimental and image-based data to evaluate the effectiveness of food waste composting as a sustainable resource recovery solution.
Assist in organizing, labeling, and managing image datasets for machine learning model development.
Assist in developing and testing machine learning or computer vision models for image-based analysis of food waste composition and fertilizer quality.
Analyze experimental and image-based data to evaluate the effectiveness of food waste composting as a sustainable resource recovery solution.
Applicant's Learning Objectives
Apply image analysis and machine learning techniques to characterize food waste and composted fertilizer.
Learn how to collect, label, and process image datasets for AI model development.
Analyze experimental data and AI model outputs to evaluate the effectiveness of food waste composting as a sustainable resource recovery solution.
Learn how to collect, label, and process image datasets for AI model development.
Analyze experimental data and AI model outputs to evaluate the effectiveness of food waste composting as a sustainable resource recovery solution.
Complexity of the project
Moderate