Design of Radial Basis Function Neural Networks with Principal Component Analysis and Linear Discriminant Analysis for Black Plastic Identification

Due to the environmental cost increased by the plastic waste, recycling of plastic waste is considered as an alternative to landfills. In order to recycle and reuse the plastic waste, it is very important to identify plastic wastes in terms of their resins. We use near infrared radiation with attenuated total reflectance and Raman spectroscopy equipment to obtain the spectra of the plastic wastes. Radial Basis Function Neural Networks are used to classify the plastic materials into several categories based on the obtained spectra. Principal Component Analysis and Linear Discriminant Analysis techniques are used to reduce the dimension of the input spectra.