CD Skripsi

Klasifikasi Sampah Plastik Menggunakan Pencitraan Komputer Dan Metode Deteksi Objek

---

ABSTRACT
Plastic waste is one of environmental problem which needs immediate solution. The amount of plastic waste in Indonesia has reached 5.63 million tons per year or 17.1% of the total national waste in 2021. Waste segregation is still operated manually and traditionally. Digital waste segregation or separation using imaging methods is a potential method to obtain efficient and low cost recycling process. Waste management is more efficient by implementing a classification method using computer vision and object detection, in this case using YOLO (You Only Look Once) model. This study used computer vision and YOLO to analyze the external features of plastic waste and classify it into 4 classes as transparent plastic bottle, color plastic bottle, transparent plastic cup, and color plastic cup. In this study, the imaging system consisted of a color camera, a conveyor, and a YOLO-based detection program. The detection algorithm had a bounding box that displayed the accuracy percentage and type of the object. The results of shows that the percentage accuracy of transparent plastic bottle was 100.00%, color plastic bottle was 88.00%, transparent plastic cup was 84.00%, and color plastic cup was 98.00%. Validation of the classification process system agrees using confusion matrix showed 98.63% for transparent plastic bottle, 98.54% for color plastic bottle, 94.44% for transparent plastic cup, and 96.92% for color plastic cup.
Keywords: Computer Vision, Plastic Waste, Object Detection Method, Classification, Plastic types

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

1803111828

Penerbit

Pekanbaru : Universitas Riau - FISIP -FISIKA., 2023

Deskripsi Fisik

-

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

1803111828

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

-

Subjek

FISIKA

Info Detail Spesifik

-

Pernyataan Tanggungjawab

tethi

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar