High-Level Project Summary
Plastic pollution is Its effects are not limited to the environment and water, but also animals' wellbeings, Regarding Nasa's articles inspired us to create a simple and effective solution. The idea of our project highlights the use of the Raspberry Pi device to monitor marine debris on the beach. In addition, we utilized artificial intelligence identify and classify marine debris and present an analysis of the results found. Developed a high-quality system that can detect plastic in the sea using massive datasets to give good results.Furthermore, we work on the latest most popular Object detection mechanisms that solve the problem with a few costs.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
We are the "vision of us" team.
Our aim Finding Plastic Patches in Coastal Waters using both NASA open-source[1], Deep Plastic(Deep Trash YOLOv5)datasets.[5]
-Yolov5 has been used as it one of the most popular Object detection mechanism.
-Raspberry Pi is low-cost and highly scalable in sensor type and node count, making it suitable for a wide range of environmental monitoring applications.[2]
-A database was created based on artificial intelligence that helps detect plastic with reasonable accuracy. A Raspberry Pi was used to build the system, a computer the size of a hand, and cameras were placed on it to identify and detect plastic. This device is placed in our places on the coasts
In order to create the system, we used 4k pictures and CoLab pro instead of another GPU.
We trained YOLOv5 on 500 epochs with a batch size of 128 using the Deep Trash YOLOv5dataset.
We got these results
These outcomes can still be improved with more instruction.
We are currently suggesting this, but we hope in the future the monitoring will be carried out by satellites owned by King Abdulaziz City for Science and Technology.
We also suggest that our project will be of benefit to the National Center for Environmental Compliance in the Kingdom of Saudi Arabia. These devices are installed on beaches in various cities of the Kingdom, and Plastic Marine Debris is monitored through these devices. Moreover, thus causing pollution, a sound warning and message are sent to the environment monitors and inspectors.It can be used for other purposes, for example, connecting it to a rescue device that automatically cleans and removes Plastic Marine Debris by itself.
We can launch the "Sea without Plastic" initiative by taking advantage of our project. By installing our system on a robot and letting children or beach walkers participate with the robot and the observers in detecting and identifying the locations of debris and marine plastic waste locations to raise their awareness of the danger of pollution on Earth.
Acknowledgements
We are grateful for Dr Saleh Albelwi, The Industrial Innovation and Robotics Center (IIRC) Director, chairman of the computer science department at the University of Tabuk, sbalawi@ut.edu.sa. we appreciate his help and guidance because he loaned us the Raspberry Pi and its accessories from the centre and helped us operate it and install the necessary libraries.
Space Agency Data
Due to time constraints, we are currently trying to provide a clear idea and understanding of NASA data on plastic litter, plastic pollution, marine plastic pollution and the identification and monitoring of marine plastic waste. Also, we found a dataset classifying the type of plastic.
The photos below demonstrated the finding of our research.
Utilize the CYGNSS Level 2 Science Data Record Version 2.1 dataset to investigate the microplastic waste that contributes about 80%-85% of all marine litter. Research published examining this dataset found a correlation between the behaviour of the wind and the spread of microplastic across the oceans' surface.
This figure below refers to the global distribution and ocean MSS (mean square slope) anomaly observations, two microplastic concentrations (#/km2, log10 scale) predicted by the van Sebille model, and three retrieved microplastic concentrations from the observations. When MSS anomalies hit near zero, we noticed that there’s a low retrieved microplastic concentration. While this dataset was collected from satellite images, we believe that closer to the surface data-collecting approach can help strengthen the results of the models used.
Hackathon Journey
Pleasant experience, the time was concise, but it was greatly benefited.
Benefits:
The ability to defy the odds in a short time. He also collected several ideas from group members in one idea and worked on them to develop the project.
Difficulties:
The lack of a GPU on our devices and the performance was slow, but Colab Pro solved the problem to speed up the training process in a short time.
Inspired your team to choose this challenge:
As for the importance of education in our specialization in technology, we have harnessed this technology to contribute to preserving the seas and oceans from environmental pollution. We are trying to reduce the volume of garbage that enters the seas and oceans.
References
1. EOSDIS Ocean Data, EARTHDATA Open Access For Open Science NASA, MAY. 25, 2021. Accessed on: OCT. 01, 2021. [Online] https://earthdata.nasa.gov/learn/discipline/ocean
2. S. G. Nikhade, “Wireless sensor network system using Raspberry Pi and zigbee for environmental monitoring applications,” in 2015 International Conference on Smart Technologies and Management for Computing, Communication, Controls, Energy and Materials (ICSTM), 2015, pp. 376–381, doi: 10.1109/ICSTM.2015.7225445.
3. Jonathan Amos Science correspondent, Earth is becoming 'Planet Plastic', BBC NEWS, July. 19, 2017. Accessed on: OCT. 01, 2021. [Online] https://www.bbc.com/news/science-environment-40654915.
4. R. Geyer, J. R. Jambeck, and K. L. Law, “Production, use, and fate of all plastics ever made,” Science advances, vol. 3, no. 7, pp. e1700782–e1700782, 2017, doi: 10.1126/sciadv.1700782
5. Gautam T. and Sarah-Jeanne R. and Olivier P. and Jay L.,(2021) DeepPlastic: A Novel Approach to Detecting Epipelagic Bound Plastic Using Deep Visual Models[Source code]. https://github.com/gautamtata/DeepPlastic
6. Roboflow-Train-YOLOv5, Colab,Accessed on: OCT. 01, 2021. [Online], https://colab.research.google.com/drive/1gDZ2xcTOgR39tGGs-EZ6i3RTs16wmzZQ
7. Ballent, Al. , Toxicological Threats of PlasticMant. EPA., May 10, 2021 , Accessed on: Oct. 02, 2021. [online] https://www.epa.gov/trash-free-waters/toxicological-threats-plastic.
8. Madeline C. Evans, Toward the Detection and Imaging of Ocean Microplastics With a Spaceborne Radar, 09 June, 2021, Accessed on: Oct. 02, 2021. https://ieeexplore.ieee.org/document/9449485?denied=
National Geographic. "Medieval helpdesk with English subtitlesSee How It Feels to Be an Ocean Animal Stuck in a Plastic Bag," YouTube, Jun 8, 2016 [Video file]. Available:https://youtu.be/yaDx-WJAsaE . [Accessed: OCT. 02, 2021].
Tags
#vision_ofus_space
Global Judging
This project has been submitted for consideration during the Judging process.