Abstract. This study investigates the use of a hybrid neural network architecture that combines convolutional neural networks (CNN) and long short-term memory (LSTM) for efficient sorting of plastic containers. The study focuses on classifying plastic waste based on chemical composition and contamination level, obtained with a near-infrared (NIR) spectroscopy device. Experimental results show that the CNN+LSTM hybrid model achieves relatively high accuracy in recognizing different types and colors of plastics, including the detection of contaminants in containers. A comparative evaluation of the model’s performance was conducted with traditional classification methods such as logistic regression, partial least squares (PLS), and linear discriminant analysis (LDA). The results show that the CNN+LSTM model performs better than traditional approaches, especially in scenarios with small spectral differences between classes. This study demonstrates the potential of machine learning to improve the efficiency of plastic waste sorting and recycling processes, thereby contributing to improved environmental sustainability.

Keywords. Plastic waste, NIRS (Near-infrared spectroscopy), Neural network, Hybrid model, CNN, LSTM.