PENGEMBANGAN SISTEM IDENTIFIKASI DAN KLASIFIKASI JENIS BOTOL PLASTIK DENGAN METODE JARINGAN SARAF TIRUAN
Daftar Isi:
- The problem that occurs in the plastic waste sorting system is the process of identifying and classifying the plastic waste. This research aims to develop an identification and classification system for plastic bottles that utilizes an intelligent computer vision system for the process of sorting plastic bottles. Development of the identification and classification system for plastic bottles using Matlab R2012b software with Backpropagation Neural Network method integrated with webcam, Red, Green and Blue (RGB) color space and lighting using blue, red and yellow lights for each image capture. Database input in the form of red, green and blue values contained in a plastic bottle object. The output is in the form of PET, HDPE and PP plastic types. The development of identification and classification systems can separate plastics based on their types namely PET, HDPE and PP automatically developed successfully. Database development and design through an image training process with 90 plastic bottle samples based on color characteristics using three blue, red and yellow lighting each obtained an accuracy percentage of 43.33%, 42.96% and 45.92 &. The percentage of successful testing using 50 samples of plastic bottles in the blue, red and yellow lights of this study was 18%, 28% and 42%.