This study contributes and focuses on identifying trash materials found in Pasig River, mainly Manila City from Santa Ana to Intramuros. The proponents will use Scaled-YOLOv4, which is a state-of-the-art object detection model that will be trained to specifically detect plastic and paper.
The main concern of this study is to test the performance of Scaled-YOLOv4-CSP configuration in detecting paper and plastic found on the surface of Pasig River, Philippines. Specifically, the study aims to answer the following questions:
1. How will the dilapidated trash dataset affect the performance of the Scaled-YOLOv4 model in detecting objects?
2. What will the Scaled-YOLOv4 model’s two classes yield in terms of precision?
3. How will the environment hinder the Scaled-YOLOv4 model in detecting objects?
In Pasig City, Philippines there is a river ferry service which is the only water-based transportation in Metro Manila that spans from Pinagbuhatan Pasig City, Mandaluyong City, Makati City, and Intramuros Manila City. It is owned by a private company called SCC Nautical Transport Services Incorporated. It is more similar to a water taxi than a ferry, and other water vessels also use the Pasig River as a route for transportation.
In previous years, Pasig river is known for being one of the most polluted rivers in the Philippines. Trashes in the water surface cause major damages to water vessels passing through bodies of water. The damage done to these water vessels cost a lot of money in repairing it. The river is now clean due to the recent efforts of the Pasig River Rehabilitation Commission. Having a clean river doesn’t mean that it will stay like that forever, a viable solution is having an automated trash detection on the surface of the water. These can be a way to minimize water debris and pollution. The detected trash can be documented on a weekly basis and notify the local government if there's an alarming increase in trash on the surface of the water.
Table of Contents
- Introduction
- Statement of the Problem
- Objectives of the Study
- Significance of the Study
- Scope and Limitation
- Related Works
- Deep Learning
- YOLO (You Only Look Once)
- Real Time Object Detection
- Damages to Boats caused by Water Pollution
- Automated Waste Sorting
- Methodology
- Machine Specifications
- Theoretical Framework
- Conceptual Framework
Objectives and Key Themes
This study aims to evaluate the performance of the Scaled-YOLOv4 object detection model in identifying plastic and paper trash on the surface of the Pasig River in Manila, Philippines. The research utilizes a dilapidated trash dataset to assess the model's robustness in challenging conditions.
- Performance of Scaled-YOLOv4 with a dilapidated trash dataset.
- Precision of Scaled-YOLOv4 in detecting plastic and paper trash.
- Impact of environmental factors on the model's performance.
- Application of Scaled-YOLOv4 for automated trash detection in a real-world setting.
- Analysis of existing object detection methods for water surface trash.
Chapter Summaries
Introduction: This chapter introduces the context of the study, focusing on the pollution problem in the Pasig River and the need for automated trash detection. It highlights the challenges of detecting dilapidated trash due to its irregular shapes and the use of Scaled-YOLOv4 as a suitable object detection model for this task. The chapter sets the stage by emphasizing the significance of the research in addressing environmental concerns and contributing to improved river management.
Related Works: This chapter reviews existing literature on relevant topics including deep learning applications in waste detection and sorting, the YOLO family of object detection models, real-time object detection techniques, and the damage caused by water pollution to boats. It establishes the groundwork for the current research by comparing and contrasting the proposed method with previous approaches and identifying gaps in the existing literature. The review encompasses a broad range of studies, demonstrating a comprehensive understanding of the field.
Methodology: This chapter details the methodology employed in the study. It describes the hardware and software used (Google Colaboratory PRO with a NVIDIA Tesla P100 GPU), the theoretical framework supporting the use of Scaled-YOLOv4, and the conceptual framework explaining the model's architecture and workflow. This section provides a clear and concise explanation of the experimental setup, allowing for reproducibility and assessment of the study's validity.
Keywords
Scaled-YOLOv4, Computer Vision, Object Detection, Trash Detection, Pasig River, Dilapidated Trash Dataset, Plastic, Paper, Average Precision, Environmental Factors, Deep Learning.
Frequently Asked Questions: Scaled-YOLOv4 for Dilapidated Trash Detection in the Pasig River
What is the main focus of this study?
This study evaluates the performance of the Scaled-YOLOv4 object detection model in identifying plastic and paper trash on the surface of the Pasig River in Manila, Philippines, specifically focusing on its ability to handle "dilapidated" or irregularly shaped trash.
What are the key objectives of the research?
The research aims to assess the performance and precision of Scaled-YOLOv4 in detecting plastic and paper trash, understand the impact of environmental factors on the model's accuracy, and explore the applicability of Scaled-YOLOv4 for automated trash detection in real-world scenarios. It also analyzes existing object detection methods for water surface trash.
What dataset was used in this study?
The study utilizes a "dilapidated trash dataset," meaning a dataset of images containing trash items that are damaged, irregular in shape, and potentially difficult to detect compared to neatly shaped trash.
What is the significance of using a dilapidated trash dataset?
Using a dilapidated trash dataset allows for a more realistic evaluation of the model's robustness and generalizability in real-world conditions, where trash is often not neatly arranged or in perfect condition.
Which object detection model is used in this study?
The study employs the Scaled-YOLOv4 model, a variation of the popular YOLO (You Only Look Once) family of object detectors known for its speed and accuracy.
What are the key themes explored in this research?
Key themes include the performance of Scaled-YOLOv4 with challenging datasets, the precision of the model in specific trash detection, the impact of environmental conditions on detection accuracy, and the practical application of the model for automated trash removal.
What is the methodology employed in the study?
The methodology involves using Google Colaboratory PRO with an NVIDIA Tesla P100 GPU. The study describes both the theoretical framework supporting the use of Scaled-YOLOv4 and the conceptual framework outlining the model's architecture and workflow.
What are the chapter summaries included in the preview?
The preview includes summaries of the Introduction (setting the context and highlighting the research problem), Related Works (reviewing existing literature on related topics), and Methodology (detailing the experimental setup and resources used).
What are the keywords associated with this research?
Keywords include Scaled-YOLOv4, Computer Vision, Object Detection, Trash Detection, Pasig River, Dilapidated Trash Dataset, Plastic, Paper, Average Precision, Environmental Factors, and Deep Learning.
What is the overall goal of this research concerning the Pasig River?
The ultimate goal is to contribute to improved river management and address environmental concerns by developing and evaluating an effective method for automated trash detection in the polluted Pasig River.
Where can I find more information on related works?
The "Related Works" chapter provides a comprehensive review of existing literature on deep learning applications in waste detection, YOLO object detection models, real-time object detection techniques, and the impacts of water pollution.
- Quote paper
- Timothy Kyle Chan (Author), 2021, Trash Detection for Computer Vision. Using Scaled-YOLOv4 on Water Surface, Munich, GRIN Verlag, https://www.grin.com/document/1151588
