.png){kind=icon}
.png){kind=icon}
Return
Rapid type identification of plastic waste using valley-side slope of near-infrared spectrum
.png){kind=icon}
PRE
AI
Z
X
L
G
DOI：10.2306/scienceasia1513-1874.2026.033.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The identification and sorting of different types of plastic waste are of great significance for improving the efficiency and quality of recycling, which in turn avoids plastic waste pollution, reduces greenhouse gas emissions, and saves production resources. The existing identification method based on spectral analysis faces the problems of complex spectral processing, high data computation, and a slow prediction speed of the prediction model, which in turn leads to low identification efficiency. To solve these problems, a rapid plastic waste type identification method using the valley-side slope of the near-infrared spectrum was developed. Three wavelengths at the left side, center, and right side of the valley and their corresponding spectral reflectance were used to calculate the left and right valley-side slopes, respectively, to construct a valley-side slope feature. Based on the constructed valley-side slope feature, identification models realizing two-category, three-category, four-category, five-category, six-category, and seven-category identification modes were established using the Classification and Regression Tree (CART) algorithm. The ten-fold cross-validation results showed that the identification accuracies of the above identification modes reached 98.4%-100%, and the prediction speeds reached 31000-39000 observations/second. The proposed method can simplify spectral data processing, reduce model complexity, and improve prediction speed, while ensuring high identification accuracy.
Keywords:
plastic waste
rapid type identification
near-infrared spectrum
valley-side slope
classification and regression tree
Journal
S
IF:
0.6
Papers:
75
Citations:
969
Organization
Cited Papers