Table 1.
Literature review on material classification and spectroscopic data analysis.
Fig 1.
The block diagram of the proposed methodology utilized for the classification of three stage PET martials.
Fig 2.
Dataset samples.
Fig 3.
Raw spectra data.
Fig 4.
Mean NIR spectra of different plastic bottle materials.
Fig 5.
Mean NIR spectra of PET bottles with different transparency levels.
Fig 6.
Mean NIR spectra of clean and contaminated PET bottles.
Table 2.
Notations and definitions used in the algorithm.
Fig 7.
The architecture of the proposed HAttFFNN model.
Fig 8.
Training curves of proposed HAttFFNN model for Stage 1: PET Clear Vs PET Hazard.
Fig 9.
Training curves of HAttFFNN model for Stage 2: PET Vs Others.
Fig 10.
Training curves of proposed HAttFFNN model for stage 3: PET Coloured Vs PET Transparent.
Table 3.
Comparative training analysis of the proposed model along with state-of-the-art methods.
Table 4.
Performance metric of the proposed HAttFFNN for stage 1: PET Clear vs PET Hazard.
Fig 11.
Confusion matrix of proposed HAttFFNN model for Stage 1: PET Clear Vs PET Hazard.
Table 5.
Performance metric of proposed HAttFFNN model for stage 2: PET vs Others.
Fig 12.
Confusion matrix of proposed HAttFFNN model for stage 2: PET Vs Others.
Table 6.
Performance metric of proposed HAttFFNN model for stage 3: PET Coloured vs PET Transparent.
Table 7.
Comparative testing analysis of the proposed model along with state-of-the-art methods.
Fig 13.
Confusion matrix of proposed HAttFFNN model for stage 3: PET Coloured Vs PET Transparent.
Table 8.
Statistical comparison of accuracy between the proposed model and existing model using T-Trest and Wilcoxon Signed-Rank Test.
Table 9.
Time complexity comparison of proposed model with state-of-the-art methods across all classification stages.