FADHLURRAHMAN, RAKHA ARDANTA (2026) PERANCANGAN SISTEM KLASIFIKASI SAMPAH PADA TEMPAT SAMPAH PINTAR MENGGUNAKAN COMPUTER VISION. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (473kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
BAB 1.pdf Restricted to Registered users only Download (176kB) |
|
|
Text (BAB II)
BAB 2.pdf Restricted to Registered users only Download (300kB) |
|
|
Text (BAB III)
BAB 3.pdf Restricted to Registered users only Download (339kB) |
|
|
Text (BAB IV)
BAB 4.pdf Restricted to Registered users only Download (571kB) |
|
|
Text (BAB V)
BAB 5.pdf Restricted to Registered users only Download (111kB) |
|
|
Text (DAFTAR PUSTAKA)
Daftar Pustaka.pdf Restricted to Registered users only Download (175kB) |
|
|
Text (LAMPIRAN)
Lampiran.pdf Restricted to Registered users only Download (125kB) |
Abstract
The increasing amount of waste requires the application of technology to support a more effective and automated waste sorting process. Manual waste sorting still has several limitations, including improper waste placement according to waste type and low sorting efficiency. This study aims to design and implement a waste classification and sorting system for a smart trash bin using Computer Vision technology. The system uses a webcam as an image acquisition device, Google Teachable Machine as a platform for training the classification model, Python and Keras for the inference process, an Arduino UNO as the controller, and servo motors as actuators for the sorting mechanism. The model was trained to recognize eight classes, namely batteries, eggs, food waste, glass waste, leaves, paper waste, plastic bottles, and the absence of waste. The classification results were then grouped into three main categories: organic waste, non-organic waste, and hazardous waste, which were subsequently transmitted to the Arduino UNO to control the sorting mechanism in real time. The system was tested through 20 trials for each waste object. Based on the classification test results, the overall classification success rate reached 81.43%, with the highest classification success rate achieved for eggs and glass waste at 100%, while plastic bottles had the lowest success rate at 40%. The model produced a confidence score of up to 100%, while the fastest average response time was approximately 1 second for plastic bottle objects. The sorting test showed an overall reliability rate of 81.67% across all tested waste categories. The results indicate that the developed system is capable of performing waste classification and sorting automatically in real time. However, the system's performance remains subject to limitations influenced by the visual characteristics of the objects and lighting conditions. Keywords: Computer Vision, Google Teachable Machine, Waste Classification, Arduino UNO, Smart Trash Bin Permasalahan sampah yang terus meningkat memerlukan penerapan teknologi yang dapat membantu proses pemilahan secara lebih efektif dan otomatis. Pemilahan sampah secara manual masih memiliki keterbatasan, antara lain ketidaksesuaian penempatan sampah berdasarkan jenisnya serta rendahnya efisiensi proses pemilahan. Penelitian ini bertujuan untuk merancang dan mengimplementasikan sistem klasifikasi dan pemilahan sampah pada tempat sampah pintar menggunakan teknologi Computer Vision. Sistem menggunakan webcam sebagai perangkat akuisisi citra, Google Teachable Machine sebagai platform pelatihan model klasifikasi, Python dan Keras untuk proses inferensi, Arduino UNO sebagai pengendali, serta motor servo sebagai aktuator mekanisme pemilahan. Model dilatih untuk mengenali delapan kelas, yaitu baterai, telur, sampah makanan, sampah kaca, daun, sampah kertas, botol plastik, dan kondisi tidak ada sampah. Hasil klasifikasi kemudian dikelompokkan menjadi tiga kategori utama, yaitu sampah organik, non organik, dan bahan berbahaya, yang selanjutnya dikirimkan ke Arduino UNO untuk mengendalikan mekanisme pemilahan secara real time. Pengujian klasifikasi dilakukan sebanyak 20 kali percobaan pada masing-masing objek sampah. Berdasarkan hasil pengujian klasifikasi, tingkat keberhasilan klasifikasi mencapai 81,43%, dengan tingkat keberhasilan tertinggi pada sampah telur dan sampah kaca sebesar 100%, sedangkan tingkat keberhasilan terendah terdapat pada botol plastik sebesar 40%. Model menghasilkan confidence score hingga 100%, sedangkan waktu respons tercepat memiliki rata-rata sekitar 1 detik pada objek botol plastik. Pengujian pemilahan menunjukkan tingkat keandalan sebesar 81,67% berdasarkan keseluruhan kategori sampah yang diuji. Hasil penelitian menunjukkan bahwa sistem yang dirancang mampu melakukan klasifikasi dan pemilahan sampah secara otomatis dan real time, namun masih terdapat keterbatasan terhadap performa sistem, dimana dipengaruhi oleh karakteristik visual objek, dan kondisi pencahayaan. Kata Kunci : Computer Vision, Google Teachable Machine, Klasifikasi Sampah, Arduino UNO, Tempat Sampah Pintar
Actions (login required)
 |
View Item |
