Date of Publication

5-19-2022

Document Type

Bachelor's Thesis

Degree Name

Bachelor of Science in Computer Engineering

Subject Categories

Computer Engineering

College

Gokongwei College of Engineering

Department/Unit

Electronics And Communications Engg

Thesis Advisor

John Anthony C. Jose

Defense Panel Chair

Roderick Y. Yap

Defense Panel Member

Maria Antonette C. Roque
Melvin K. Cabatuan

Abstract/Summary

Improper disposal of waste can lead to environmental problems such as pollution that negatively impacts the ecosystem. Proper waste segregation is an essential practice to help maintain the sustainability of the environment. This results in a significant reduction in the number of garbage to be disposed of in landfills. This research aims to develop an automated waste segregator that will classify and sort waste objects according to their group (biodegradable, non-biodegradable, and recyclable). With the help of a CNN model that is trained using transfer learning approaches, computer vision technology made it possible to identify waste objects. 15 waste object classes were specified , 5 objects per group. It was found that the performance of the CNN model is highly dependent on the quality of the dataset and the source model of choice. With experiments, our best performing CNN model was obtained by training a total of 8,621 images from 15 different classes with ResNet-50 as our source model.

Abstract Format

html

Language

English

Format

Electronic

Keywords

Sorting devices; Refuse and refuse disposal

Upload Full Text

wf_yes

Embargo Period

6-25-2022

Download

Share

COinS