CD Skripsi
Analisa Karakteristik Spektrum Sampah Plastik Dan Sampah Organik Menggunakan Pencitraan Multispektral
ABSTRACT
Waste is still a crucial problems faced by global society daily. It requires further waste management. The waste sorting system has implemented an imaging method for a plastic and organic waste sorting system that is effective in identifying and distinguishing waste based on color or chemical substances. Waste segregation is important to facilitate the recycling process of polyethylene terephthalate (PET), high density polyethylene (HDPE), polypropylene (PP) plastic and organic waste before reaching the landfills. This research aims to use a multispectral imaging system to identify and characterize the reflectance intensity of plastic and organic waste. The multispectral imaging system uses a filter wheel with 10 bandpass filters that has several wavelangths such as 520, 680, 710, 740, 770, 800, 830, 860, 880 and 890 nm. The results showed a significant difference in the relative reflactance intensity of multispectral images, especially at a wavelangth of 710 nm. HDPE plastic has the highest intensity because many samples was found to be colored and thick. The peak relative reflectance intensity of multispectral images is higher on paper and cardboard than on wet organic wastes. Principal component analysis (PCA) method was used to prosess data in order to facilitate data visualization. Two variables ware used to analyze multispectral images, namely PC1 and PC2 with a cumulative percentage of 94,56% for plastic waste and 97,37% for organic waste. The pattern formed on the scatter plot show that there is a grouping of data in each type of sample according to the characteristics. These finding will be use for building an Artificial Neural Network (ANN).
Keywords : Plastic Waste, Organic waste, Multispectral Imaging, Principal Component Analysis, wavelangth dependence
Tidak tersedia versi lain