Related Experiment Video
Updated: Jul 22, 2025
10:12
Extraction of Organochlorine Pesticides from Plastic Pellets and Plastic Type Analysis
Published on: July 1, 2017
11.6K
Multi-level color classification of post-consumer plastic packaging flakes by hyperspectral imaging for optimizing
Paola Cucuzza1, Silvia Serranti1, Giuseppe Capobianco1
1Department of Chemical Engineering, Materials & Environment, Sapienza University of Rome, Rome, Italy.
Summary
This study introduces a hyperspectral imaging and machine learning method for sorting high-density polyethylene (HDPE) plastic waste by color. This approach enhances the quality of recycled HDPE, supporting circular economy principles.
Area of Science:
- Materials Science
- Computer Science
- Environmental Science
Background:
- Effective plastic waste sorting is crucial for high-quality secondary raw material production in recycling.
- Current methods require efficient strategies for recognizing plastic waste by both polymer type and color.
Purpose of the Study:
- To develop and compare sensor-based recognition strategies for classifying mixed-colored high-density polyethylene (HDPE) flakes by color.
- To optimize plastic recycling processes for a circular economy.
Main Methods:
- Utilized hyperspectral imaging in the visible range (400-750 nm) combined with machine learning.
- Developed two classification models: Partial Least Square-Discriminant Analysis (PLS-DA) for 6 macro-color classes and hierarchical PLS-DA for 14 color tones.
Main Results:
- Both classification models achieved excellent performance, with prediction metrics (Recall, Specificity, Accuracy, F-score) close to 1.
- The hierarchical PLS-DA model provided more accurate discrimination of HDPE color tones.
Conclusions:
- The proposed sensor-based sorting approach is effective for plastic recycling plants.
- This method enables the production of high-quality recycled HDPE in various colors, aligning with circular economy principles.
