Rancang Bangun Sistem Pemilah Jenis Sampah Otomatis Berbasis Image Processing

Sabila, Lailiyatul Fharichatus (2026) Rancang Bangun Sistem Pemilah Jenis Sampah Otomatis Berbasis Image Processing. Diploma thesis, Institut Teknologi Sepuluh Nopember.

Text 2042221113-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Metode pemilahan sampah konvensional yang mencampur berbagai jenis limbah sejak dari sumber seringkali menghambat efisiensi proses daur ulang. Penelitian ini bertujuan merancang dan membangun sistem pemilah jenis sampah otomatis berbasis image processing. Sistem ini memanfaatkan kamera webcam dan algoritma Convolutional Neural Network (CNN) dengan arsitektur MobileNetV3Small yang dijalankan pada Raspberry Pi 5 untuk mengklasifikasikan sampah ke dalam kategori organik, anorganik, dan logam. Sistem dilengkapi dengan antarmuka GUI untuk memantau proses klasifikasi dan jumlah sampah yang terpilah secara real-time. Hasil pengujian model menggunakan 2.061 citra latih menunjukkan akurasi sebesar 97,10% dengan waktu inferensi 40 ms. Pengujian langsung (real-time) terhadap 574 sampel sampah menunjukkan tingkat akurasi sistem sebesar 97,91% dengan tingkat kesalahan (error rate) sebesar 2,09%.
===========================================================================================================================
Conventional waste sorting methods that mix various types of waste from the source often hinder the efficiency of the recycling process. This study aims to design and build an automatic waste sorting system based on image processing. This system utilizes a webcam camera and a Convolutional Neural Network (CNN) algorithm with MobileNetV3Small architecture running on a Raspberry Pi 5 to classify waste into organic, inorganic, and metal categories. This system is equipped with a GUI interface to monitor the classification process and the amount of waste sorted in real-time. The results of model testing using 2,061 training images showed an accuracy of 97.10% with an inference time of 40 ms. Direct (real-time) testing on 574 waste samples showed a system accuracy level of 97.91% with an error rate of 2.09%.

Item Type:	Thesis (Diploma)
Uncontrolled Keywords:	MobilenetV3Small, CNN, klasifikasi sampah, akuisisi citra, transfer learning, garbage classification, image acquisition
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1573 Detectors. Sensors T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing. T Technology > TA Engineering (General). Civil engineering (General) > TA593.35 Instruments, cameras, etc.
Divisions:	Faculty of Vocational > Instrumentation Engineering
Depositing User:	Lailiyatul Fharichatus Sabila
Date Deposited:	05 Aug 2026 03:46
Last Modified:	05 Aug 2026 03:46
URI:	http://repository.its.ac.id/id/eprint/141057

Actions (login required)

View Item